A number of trials of adoptive transfer of tumor-specific T lymphocytes have been performed in the last 20 years in metastatic melanoma, with increasingly encouraging results as the relevant melanoma antigens were identified and the purity/specificity of injected T cells improved. We have previously described a sorting method of epitope-specific T lymphocytes that uses magnetic beads coated with HLA/peptide complexes and we suggested that this method could be applied to a clinical setting. In the present work, we provide a detailed description of the whole GMP process of sorting and amplification of clinical grade T cells specific for the melanoma antigens Melan-A and MELOE-1. All the reagents used in this process including the sorting reagent were produced in GMP conditions and we document the optimization of the different steps of the process such as peptide stimulation, sorting, and amplification. The optimized procedure, validated in 3 blank runs in a clinical setting, allowed the production of at least 108 pure (>90%) Melan-A- and MELOE-1-specific T cells within 28 days starting with 100 mL of blood from metastatic melanoma patients. This GMP process is thus ready to be used in an upcoming phase I/II clinical trial on metastatic melanoma patients.
In cancer, the best argument in favor of adoptive cell transfer (ACT) is the demonstration that it can elicit clinical regressions of cancers not curable by other treatments. Initially established for hematopoietic tumors in an allogeneic setting, the beneficial effect of ACT has also been documented in autologous situations such as the control of EBV-induced tumors by virus-antigen-specific T cells [
To select and expand such specific T cells, we developed an original method of T cell sorting from blood sample, based on the selection of specific T cells with HLA-peptide coated magnetic beads [
To adapt this procedure for a clinical use, this requires that the cell production facility, the ancillary products, and the production process used meet the GMP (Good Manufacturing Practices) criteria in terms of quality controls (characterization and purity of produced T cells), safety (absence of microbiological contamination), and robustness (reproducibility and repeatability of the production process).
To this aim, in collaboration with an industrial partner, PX’Therapeutics, we produced clinical grade batches of HLA-peptide-coated magnetic beads (Clinimers) in order to treat seventeen HLA-A2 metastatic melanoma patients at stage IIIc (not operable lymph node metastases) or IV (distant metastases) in an upcoming clinical trial. Starting from 100 mL of blood, Melan-A- and MELOE-1-specific T cells will be selected and amplified
Blood was collected from healthy HLA-A2 donors (Etablissement Français du Sang (EFS), Nantes, France) or from metastatic melanoma patients (Unit of Skin Cancer, Centre Hospitalier Universitaire Hotel Dieu, Nantes) after written informed consent.
The two melanoma cell lines M113 and M117 were established from metastatic tumor fragments in the Unit of Cell therapy of Nantes and are registered in the Biocollection PC-U892-NL (CHU Nantes).
Peripheral blood mononuclear cells (PBMC) were isolated by Ficoll-Hypaque gradient centrifugation, washed three times, and seeded in 96 well/plates at 2 × 105 cells/well in either RPMI 1640 medium supplemented with 8% human serum (HS) (a pool from 20 donors prepared and secured by the EFS of Nantes) or in X-Vivo 15 serum-free medium (Lonza, Levallois-Perret, France) with various concentrations of IL-2 (from 10 IU/mL to 150 IU/mL). PBMC were stimulated by adding various concentrations of clinical grade Melan-
The design of HLA-peptide multimers that we used for specific T cell sorting was previously described [
HLA-A0201/peptide
Final assembly of the Clinimer reagent was performed immediately before T cell sorting by incubating Chim-AvT beads with the appropriate HLA/peptide monomers (1
Sorting of Melan-A and MELOE-1-specific T cells was performed by mixing a T cell suspension (5 × 106 to 107 cells) containing at least 2 × 105 specific T cells with Clinimers at a ratio of 1 bead per cell in PBS 4% HSA, as previously described [
T cell cultures are split in half every 2 to 3 days and fresh medium is added. After about 8 ± 1 days, cells are transferred to 6-well plates for a final bulk amplification of 7 days. This open culture system was chosen because of a better efficiency of T cell amplification than in dedicated cell bags (data not shown). At this step, 108 lymphocytes are seeded at 5 × 105 cells/mL in 5 mL of medium per well, containing 150 U/mL of IL-2. Every two to three days cell concentration is adjusted to 5 × 105 cells/mL in the same medium, until day 14.
The specificity of the sorted and amplified T lymphocytes was evaluated by cytokine production in response to their cognate epitope. In brief, T cells were stimulated for 5 h in the presence of brefeldin A (10
We have previously demonstrated that to ensure efficient sorting of specific T cells with HLA multimers, that is, high yields and high purity (>90%) in all donors, the starting PBMC populations should contain at least 0.5% of specific T cells and thus a short peptide stimulation is required that does not alter T cell repertoire [
Peptide stimulation step. (a) 107 PBMC from HLA-A2 healthy donors were stimulated in 96-well plated (containing 2 × 105 cells/well) for 14 days with Melan-A or MELOE-1 peptides. (b) Percentages of Melan-A- or (c) MELOE-1-specific T cells detected by double labelling with tetramer and anti-CD8 antibody, in microcultures stimulated with 0.1 to 10
For each antigen, we first defined the optimal peptide concentration for amplification of Melan-A (Figure
To favor the proliferation of specific T cells, IL-2 is added to the culture medium, during the peptide stimulation step. However, IL-2 is a double edge sword since it can also induce apoptosis by AICD [
Influence of IL-2 concentration on the efficiency of peptide stimulation. 107 PBMC from HLA-A2 healthy donors were stimulated in 96-well plated (2 × 105 cells/well) with either Melan-
The use of human serum for the amplification of T cells requires the production of dedicated batches by a transfusion center and the validation of each batch for the absence of known viruses. Because it is rather cumbersome and expensive, alternate culture conditions were tested.
Since melanoma-derived TIL are successfully amplified in X-Vivo 15 medium [
Influence of the culture medium on the efficiency of peptide stimulation. 107 PBMC from HLA-A2 healthy donors (a and b) or from melanoma patients (c) were stimulated in 96-well plated (2 × 105 cells/well) with either Melan-
As mentioned in Section
Sorting procedure. (a) 107 Chim-AvT dynabeads are coated with HLA-A2-peptide monomers, and coating efficiency is assessed on 105 beads, by labelling with an PE-labelled anti-HLA-A2 mAb (5
After the sorting step nonspecific lymphocytes are removed by successive washes on a magnet, which only retains Clinimers-coated T cells. These washes critically impact the yield and the purity of sorted T cells. Indeed, too many and/or too vigorous washes will result in Clinimers detaching from specific T cells and thus decrease the yield of the sort, while insufficient washing will lead to a lower purity due to nonspecific T cells that have not been discarded. In the sorting procedure that we initially described, 10 washes were recommended to ensure a high purity of sorted cells [
Yields were estimated by comparing the absolute number of specific T cells in the population before the sort (as measured by HLA-p tetramer staining) and the number of rosettes (Clinimers-coated T lymphocytes) obtained after the sort. These rosetted T cells are enumerated through two manual counts performed on two independent samples.
Purity of sorted cells was evaluated by HLA-p tetramer staining after amplification on feeder cells (Figure
As shown in Figures
During their polyclonal activation, sorted T cells downregulate their TCR surface expression and undergo many cell divisions. This results in rapid detachment of the Clinimers from T cells after 2 to 3 days in culture. Because the clinical grade magnetic beads used in Clinimers are not approved for injection to patients but only as an ancillary product in a cell production process, residual beads have to be removed from cell cultures, and the absence of beads has to be documented before adoptive cell transfer. This removal is performed after 8 days of amplification on cell suspensions adjusted to a concentration of 106 cells/mL in 50 mL tubes that are placed inside a circular magnet for 10 minutes. A second round of bead elimination is performed in the same conditions. A sample of each cell suspension (corresponding to 1/100 of the final volume, that is, about 100 mL at this stage) is analyzed by flow cytometry to confirm the absence of residual beads. The beads are detected by their properties of autofluorescence in each detection channel of a Facs Canto II, allowing an accurate discrimination between beads and cell debris (Figure
Detection of residual beads in amplified T cells. (a) Limit of detection of magnetic beads among a T cell population. Variable numbers of beads were mixed with 106 T cells in 1 mL. Beads were detected by their autofluorescence in the various channels of a Facs Canto II. (b) Absence of residual beads in amplified T cells. After 8 days of culture of sorted T cells, each T cell suspension is placed inside a magnet to remove magnetic beads. After two rounds of elimination, the absence of residual beads is checked by flow cytometry.
The last parameter that was evaluated was the average amplification factor of the sorted cells on irradiated feeder cells in order to calculate the initial number of PBMC needed to recover sufficient numbers of specific T cells at the end of the process. In our future clinical trial, we want to ensure the recovery of at least 108 Melan-A- and MELOE-1-specific T cells, that is, at least 2 × 108 tumor-specific T cells for injection to the patients. This minimal number of 2 × 108 antigen-specific T cells to be infused to melanoma patients was chosen after analysis of the amounts of infused T cells in previous clinical trials of adoptive transfer [
In the present study, as shown in Table
Amplification rates of antigen-specific T cells on irradiated allogeneic feeder cells.
Peptide | Number of |
Final number of |
Amplification factor1 | Average number of |
---|---|---|---|---|
Melan-AA27L | 105 |
|
1160 | 10 |
|
|
3075 | 11 | |
|
|
860 | 10 | |
|
|
2600 | 11 | |
105 |
|
1380 | 10 | |
|
|
|||
| ||||
|
|
|
2000 | 11 |
105 |
|
1970 | 11 | |
|
|
325 | 8 | |
|
|
5625 | 12 | |
|
|
900 | 10 | |
|
|
The final amplification procedure will be thus performed starting from 105 rosetted T cells, amplified on feeder cells in one 96-well plate, with 150
Purity of sorted and amplified T cells was evaluated by HLA-p tetramer/CD8 double staining. As shown in Figure
Purity and reactivity of sorted and amplified T cells. (a) Purity of sorted and amplified T cell populations is assessed by double staining with an anti-CD8 specific mAb and each specific HLA-p tetramers. (b) Activity of sorted T cells is measured by TNF production in response to the cognate peptide or to HLA-A2 melanoma cell lines expressing Melan-A and MELOE-1 antigens. After activation, T cells are stained with their specific HLA-p tetramer, and TNF producing cells are visualized by intracellular staining with an anti-TNF mAb. Blue histograms represent Melan-A- specific T cells and red ones MELOE-1-specific T cells.
Finally, the whole procedure (Figure
Validation runs for the selection and amplification of Melan-A- and MELOE-1- specific T lymphocytes from HLA-A2 melanoma patient-derived PBMC.
Run ID |
|
|
|
|
|
|
---|---|---|---|---|---|---|
Peptide | Melan-AA27L | MELOE-136-44 | ||||
Number of stimulated PBMC |
|
|
|
|
|
|
| ||||||
Number of sorted cells and fraction of specific T cells | 107 |
|
|
|
|
107 |
(9.3%) | (34%) | (38%) | (11%) | (16%) | (9%) | |
Theoretical number of specific T cells1 |
|
|
|
|
|
|
Recovered number of sorted specific T cells |
|
|
|
|
|
|
Sorting yield2 | 42% | 39% | 11% | 22% | 60% | 19% |
Final number of amplified specific T cells and amplification yield3 |
|
|
|
|
|
|
( |
( |
( |
( |
( |
( |
|
Average number of cell divisions | 15 | 13 | 16 | 14 | 10 | 15 |
Purity of amplified T cells4 | 95% | 99% | 92% | 90% | 91% | 93% |
Activity of amplified T cells on the cognate peptide5 | 82% | 71% | 72% | 75% | 53% | 50% |
Design of the whole GMP process. At day 0, 100 mL of blood is collected from HLA-A2 melanoma patients. Total PBMC are stimulated with the two antigenic peptides during 14 days. At day 14, antigen-specific T cells are sorted using Clinimers, and 105 rosetted T cells are seeded for amplification on feeder cells. At day 28, purity of amplified cells is assessed by flow cytometry and between 108 and 5 × 108 T lymphocytes specific for each peptide will be infused to the autologous melanoma patient.
This resulted in the recovery of at least 1.2 × 105 antigen-specific T cells prior to the amplification step. After a 14-day amplification period, on feeder cells, in RPMI medium supplemented with 8% HS, 150 U/mL of IL-2, and 1
The cell production process (Figure
To fulfill these criteria, we set specifications for each production step, detailed in Table
Specifications and release criteria for manufactured T cell products.
Production steps | Specifications/release criteria |
---|---|
Peptide stimulation step | ≥0.5 |
Coating of HLA-p on |
RFI ≥ 20 assessed by HLA labeling |
Sorting step | ≥105 rosetted T cells |
Amplification step | ≥108 specific T cells (RC)1 |
| |
Quality controls | |
| |
Purity | ≥90% purity assessed by tetramer labeling (RC) |
Activity | ≥50% TNF producing T cells among tetramer positive ones (RC) |
Safety controls | Absence of microbiological contamination (RC) |
If each T cell population meets these criteria, T lymphocytes are pooled and adjusted at a concentration ranging between 106 and 5 × 106 cells/mL in a volume of 200 mL of a pharmaceutical solution of 4% human albumin, in a bag which is a medical device. This cell preparation remains stable in terms of purity and activity at least 2 hours. The delay between the preparation of the cell suspension and its injection should be less than two hours. The product is then administered intravenously at a rate of 3 mL/min under medical supervision.
In conclusion, our results demonstrate the many advantages of our GMP procedure for the production of therapeutic melanoma-specific T lymphocytes. (i) From a practical point of view, the initial blood sample volume is still reasonable and can be obtained from most patients. (ii) The two selected antigenic peptides allow the expansion of specific T cells in the vast majority of HLA-A2 patients, and these T cells are reactive against melanoma cells. (iii) The short duration of the whole procedure results in the production of pure and polyclonal specific T cells that underwent a limited number of divisions and should thus persist
The therapeutic potential of the melanoma specific T cells obtained with this procedure will soon be evaluated in an upcoming clinical trial supported by the French National Cancer Institute (INCa) that will include 17 metastatic melanoma patients.
The authors declare that they have no conflict of interests.
The authors thank J. Desfrançois, from the cytometry facility “Cytocell” and the recombinant protein facility of the SFR Sante for expert technical assistance. This work was supported by a grant from the “National Research Agency,” BiotecS program 2009.