Regulatory Issues in Gene-Modified Immune Effector Cell Therapy

Regulatory Issues in Gene-Engineered T-Cells

Cancer immunotherapy using T-cells engineered to expressed T-cell receptors (TCR) or chimeric antigen receptors (CAR) has moved from investigational agents into approved treatments. Approval by the US Food and Drug Administration (FDA) requires licensed products to be both safe and efficacious. Engineered cell products transduced with integrating vectors (gammaretroviral and HIV-1-based lentiviral vectors, referred to here as retroviral and lentiviral vectors, respectively) present unique safety concerns related to the infusion of cell products and the potential adverse events associated with gene therapy. To determine the risk associated with engineered T-cells, regulators have required a challenging array of testing requirements that include analysis of the vector product, the transduced cell product, and postinfusion monitoring of patients for up to 15 years. In this chapter, we will focus on these safety issues and discuss current and proposed changes to FDA requirements related to the use of engineered T-cells.

Generating and Testing Vector Products

TCR- and CAR T-cell products are typically generated by ex vivo transduction of T-cells with a retroviral or lentiviral vector. Regulators consider the transduced cell product the investigational agent, but the unique safety concerns related to gene therapy vectors has resulted in a series of regulatory guidance documents that affect those who manufacture vector and those who utilize the vector products to modify T-cells ( Table 14.1 ).

Retroviral and lentiviral vectors are membrane-bound RNA viruses, and while HIV-1 is considerably more complex, retroviral and lentiviral vectors are similar in overall design ( Fig. 14.1A and B ). One major difference is that many retroviral vectors retain the native viral long terminal repeats (LTRs) which are required for integration and also contain the promoter and enhancer sequences used to drive transgene expression. Lentiviral vectors use self-inactivating LTRs requiring an internal promoter to drive transgene expression. As discussed below, retention of the gammaretroviral LTR has significant safety implications.

Vector products can be manufactured by two different methods ( Fig. 14.1C and D ). Traditionally, clinical retroviral vectors were generated in packaging cell lines which contained three integrated plasmids. Since the vector genome has been stripped of key viral genes required for viral replication, a plasmid expressing the gag/ pol sequences is needed to package the vector RNA into viral particles. A third plasmid encoding an envelope glycoprotein plasmid is also stably introduced into the cell line. The envelope is expressed on the virion surface and directs the particles to specific receptors on a target cell. Human cells lack the receptor for many gammaretroviruses, so alternative envelopes (such as the Gibbon Ape Leukemia Virus envelope) are used to “pseudotype” the vector particles. Packaging cells can generate vectors over many days and can be used to generate Master Cell Banks (MCBs) of the cell line, thereby providing a consistent product as the vector moves from Phase I through licensure.

The other method commonly used, particularly for lentiviral vectors, has been transient transfection ( Fig. 14.1D ). In this method the plasmids are transfected into cell lines and the vector harvested over 1–2 days. Pseudotyping is also used for most lentiviral vectors as the native HIV-1 envelope glycoprotein would limit infection to CD4+ cells. Many lentiviral vectors contain the vesicular stomatitis virus G glycoprotein (VSV-G) which provides high level of gene transfer into most mammalian cell types. An advantage of transient transfection is it bypasses the prolonged period of clone selection and MCB generation associated with packaging cell lines making it the quickest path to a Phase I trial. A possible downside is lot to lot variability of the transient method which could be an issue for later phase trials. There are now lentiviral packaging cell lines in development, and some retroviral vectors used clinically have been generated by transient transfection. The science of vector production is rapidly developing, particularly as industry seeks to meet the consistency challenges around manufacturing licensed products.

The FDA guidance documents listed in Table 14.1 provide testing requirements for cell lines used in vector manufacture along with the tests required on the vector product (see Table 14.2 ). The majority of assays listed in Table 14.2 are not unique to vector production and are required for generating any biologic product. Requirements are tailored to the cell line species of origin. If cell lines of more than one species are utilized (ex. using the “ping-pong” method to obtain high-titer packaging cell line clones ), expanded testing is required to cover pathogens from the relevant species. As shown in Table 14.2 , excluding virus contamination is a major focus of release testing for the cell line and the vector product. Since the cells used in manufacture are in fact generating viral particles, excluding replication competent retroviruses within the large number of vector particles has proven to be a major challenge for product development (discussed in detail below).

The FDA does require some specific tests when HEK293 and HEK293T-cell lines are used in vector manufacturing. Both lines contain genetic material from the adenovirus E1A region, and HEK293T-cells also contain sequences from the SV40 large T antigen. FDA requires vectors generated in these cell lines to be tested to ensure there is no passage of these viral sequences to transduced cells. Since some cellular DNA frequently contaminates vector products, evaluating transduced cells shortly after transduction will generally be positive by qPCR and other methods. Indiana University has utilized a 21-day assay, testing cells within the culture at approximately 5, 11, and 21 days. E1A or SV40 sequences in the culture decrease over time and most are negative by day 11.

Testing will also be required to show transduced cell products contain intact vector sequences. Retroviral and lentiviral vectors are prone to alternate splicing. In some cases, vectors may be generated that contain the full-length vector and an additional truncated transcript. The ratio can vary. If sequencing alone is used to analyze packaging cell lines and vector products, care must be taken to assure significant truncated transcripts can be detected. In addition to sequencing, Southern blots may be useful in documenting intact vector sequences in transduced cells. The US FDA has also required testing for residual total DNA or residual plasmid DNA, depending on the manufacturing method. Other product-specific testing may be required, such as residual benzonase in products where benzonase is used to reduce residual plasmid after transient transection. Specific limits for DNA and benzonase have not be published, and the requirement is generally to “report results” on the product Certificate of Analysis. Acceptable limits will likely be required as products move from Phase I to licensure.

Risk of RCR and RCL and Testing of Vector Products

As mentioned above, excluding replication competent retroviruses or lentivirus (RCR and RCL, respectively) has been a major concern of regulators. The design of early retroviral packaging cell lines contained significant homology between the vector and packaging sequences and RCR development during vector production was well documented. Gammaretroviruses are slowly transforming retroviruses and the Moloney murine leukemia virus (MLV), frequently used as the backbone for retroviral vectors, is known to cause lymphoma in mice. The LTR enhancer is felt to be an important factor in determining the predilection for causing lymphoma in mice, and its insertion is frequently found near the pim-1 gene. Additional mutagenic effects are required as pim-1 transgenics do not have an increased risk of malignancy, but exposure of these mice to RCR lead rapidly to lymphoma with virus insertions near c-myc and n-myc . The potential risk for humans was noted in a variety of early safety studies. Mice injected with an RCR that arose from the PA12 packaging cell line led to lymphoma that appeared similar to that caused by the parent MLV. While a large volume of RCR infused into immunocompetent, nonhuman primates did not appear to cause disease, inadvertent exposure of an immunosuppressed rhesus monkey during a bone marrow transplant experiment led to T-cell lymphoma.

Having identified malignancy as a potential risk from RCR exposure, the FDA developed a specific guidance for clinical trials using retroviral vectors entitled “Supplemental Guidance on Testing for Replication Competent Retrovirus in Retroviral Vector Based Gene Therapy Products and During Follow-up of Patients in Clinical Trials Using Retroviral Vectors.” Testing was required on 5% of the vector product and 1% of the end-of-production (EOP) cells (up to a maximum of 10 ⁸ cells). If vector is generated in a packaging cell, cell line testing is also required. The method of testing is suggested in the FDA guidance. For testing RCR in vector products, molecular assays are not sufficiently sensitive and likely to yield false-positive results due to carryover of packaging cell line DNA. Cell-based assays have been recommended and the material must be passaged on a cell line that is highly infectable and capable of expanding RCR to high titer. Since retroviral vectors have been pseudotyped with a variety of viral envelopes, a validated RCR detection method specific for the envelope must be used.

The other challenge in detecting an RCR is understanding the potential RCR genome. Most recombinant viruses arising from packaging cell lines contain sequences from the vector and packaging plasmids, but not in all cases. In mice, gammaretroviruses can recombine in vivo to form novel viruses with increased pathogenicity (reviewed in Cornetta et al. ). RCR have also been reported to contain genomic sequences from the packaging cell line used in early gene transfer experiments.

A variety of cell-based assays have been reported for the testing of retroviral and lentiviral vectors. These cell-based assays, commonly used to screen vector and cell products, are similar and the basics are depicted in Fig. 14.2 . Vector supernatant or end-of-production (EOP) cells are cultured with a cell line that is used to amplify virus to high titer. Since the assay must be capable of detecting a recombinant virus whose growth kinetics is unknown, regulators have required the extended passage (approximately 3 weeks of culture) to detect a slow-growing virus. Since the large number of vector particles could inhibit infection of the amplification cells (receptor interference), a large number of cells are required in the assay. Commercially available RCR and RCL testing can cost between $8000 and $18,000 per sample.

There is currently no public repository of testing results for retroviral products used in clinical trials. If a patient had been exposed in the United States, it would have been public knowledge based on the past reporting requirements to the US National Institute of Health Recombinant DNA Advisory Committee. To date, no such disclosure has been reported for RCR or RCL. This is likely due, in part, to improved vector and packaging cell line design. ^,

While the FDA guidance around RCR was developed prior to the use of lentiviral vector products, regulators have applied the same requirements. Lentiviral vectors by their design with self-inactivating long terminal repeats removing HIV-1 enhancers and promoters, removal of the HIV-1 accessory proteins and sequences such as Tat, frequent use of transient vector production methods, and the lack of direct malignant transformation by HIV-1 insertions predict an even safer profile for this vector class. While there is less experience with lentiviral vector products, reports have also supported the generation of RCL-free vector material using current production methods for clinical products.

Cell Product Issues