The Impact of Donor HLA Type on Characteristics of CMV-Specific T Cell Products: Experience from Two UK Randomized Trials

Objective

Investigator-led Phase I/II studies have shown that adoptive transfer of donor virus-specific cytotoxic T cell lymphocytes (CTL) can reconstitute immunity on a long-lasting basis in patients at risk for viral reactivation following allogeneic hematopoietic stem cell transplantation (allo HSCT). In an effort to provide a consistent high-purity product, it is critical to detect patterns of donor material which may predict final product quality and consequently CMV protection.

Methods

We performed two prospective studies of cytomegalovirus-specific (CMV) CTL in the UK, using a proprietary technology for direct selection of CTL based on MHC class I-multimer (Streptamer®) binding (STAGE Cell Therapeutics GmbH). Eligible donors were pre-screened for the presence of selectable CMV-specific cells by immunophenotyping with flow cytometry. Eligible HLA alleles for this selection procedure are A*0101, A*0201 (two epitopes, IE-1 and PP65), A*2402, B*0702 and B*0801. In our experience, a minimum threshold of 0.1% CD8+Streptamer+ cells of total CD3+ T cells ensures a successful selection.

The table summarizes the manufacturing data from 39 products, showing that despite a large variety of donor material, these cells can be significantly enriched to produce a viable, CMV-specific T cell product.

Results

Pre-screen results have been analyzed from 56 eligible donors. Of these donors, 40 were screened for more than one eligible HLA type. In the presence of more than one eligible allele, selection was based on the most abundant population of CD8+Streptamer+ cells.

The majority of donors screened were confirmed as eligible HLA type A02 (33/56), closely followed by A01 and B07 (both 22/56). However the alleles found to be the most dominant for Streptamer binding (giving the highest percentage of positive cells in these pre-screens) were A01 and B07 (both 11/20). Thus while A02 is a more frequent allele in the studied donor population, it is often subdominant in the presence of another Streptamer-eligible allele. Where A01 is present in combination with one other allele, A01 is more frequently dominant.

We also investigated whether the allele chosen for the selection procedure was predictive of the purity (CD3+Streptamer+) of the final CMV CTL product. We examined 39 selections. The majority of selections (22/39) were performed on A02-PP65. The selections produced a varying product with a purity range of 4.48% - 99.56% (SD 30.9%). Our data suggests that A01 and B07 are more indicative of a high purity CMV selection, however this was not demonstrated beyond statistical significance when compared to A02-PP65 (A01, p-value=0.34, B07, p-value=0.15).

Other factors were observed that influence product purity. The percentage of CD3+ cells and the total number of Streptamer positive cells in the starting material were also key variables for predicting the purity of the selection. Pre-screen results were not found to have a direct correlation with product purity. We investigated whether a successful pre-screen can be predicted on the presence of a single HLA type, regardless of other alleles present. A donor is least likely to have a selectable population for A24 or A02-IE-1. A01(median pre-screen 0.61%) and B07 (median pre-screen 0.53%) donors are more likely to have a positive pre-screen compared to the other alleles.

Conclusion

Cell Medica is developing a model for centralized production and delivery of CMV-specific CTL for transplant centers across Europe. Interrogation of historical data demonstrates important patterns for donor selection and effective pre-screen methods to ensure a consistent and high quality cell product. Further investigation is required to determine the impact of donor selection and product purity on CMV protection in patients. In addition, functional analysis of the specific cells may also be an important factor to consider in the donor and selectable epitope screening process.