Successful leukapheresis and CAR T cell manufacturing
The current recommended dose of tisagenlecleucel contains 0.2 to 5.0 x106 CAR-positive viable T cells per kg of body weight for patients 50kg or less, or 0.1 to 2.5 x108CAR-positive viable T cells for patients more than 50kg40. In order to achieve this dose, an adequate quantity of T cells must first be collected from the patient; therefore, a minimum absolute lymphocyte count (ALC) of ~500 cells/uL and a CD3+ cell count of ~150 cells/uL is recommended prior to starting apheresis 41. Factors that would affect both ALC and CD3+ cell count include timing of proximal cytotoxic therapy or progressive leukemic disease leading to bone marrow replacement of cancer cells. Some patients are never able to achieve these minimal peripheral blood parameters due to the nature of their highly refractory B-ALL. For these patients, the prospect of an allogeneic CAR T cell product, which remain in early clinical investigation, is attractive 42.
Once an adequate quantity of T cells is collected, characteristics of the leukapheresis product may directly affect the quality and/or performance of the CAR T cell product. Predicting the performance of CAR T cell products is quite difficult using in vitro testing, so at this time, performance is best assessed after infusion into the patient using the metrics of disease response, in vivo proliferation and CAR T cell persistence. Expansion is a vital element to disease response. In the ELIANA trial, expansion (measured as the geometric mean of the area under the concentration-time curve in peripheral blood from time 0 to day 28 as expressed in copies per microgram of DNA times days) was 315,000 in patients with a response and 301,000 in patients without a response21. In addition, responders to tisagenlecleucel have a shorter median time to maximum expansion of 11 days compared to 13 days in non-responding patients21,37. Much research has been dedicated to understanding the mechanisms behind poor expansion and persistence of the T cell product in order to maximize the anti-leukemic property of this drug.
First, recent studies have demonstrated that T cell phenotype plays an important role in predicting a CAR T cell product’s subsequent clinical activity. The presence of naïve and early memory T cells with significant proliferative potenital in the pre-manufactured product was found to correlate with a biomarker of successful CAR performance in pediatric B-ALL 43. Peripheral blood samples that contained a higher percentage of naïve and stem central memory cells directly correlates with T cell expansion potential in vivo44, and CAR T cell products that contain more central memory T cells persist longer, which can mediate a more successful clinical response 45,46. Interestingly, it has also been recently shown that the distribution of T cell subsets in peripheral blood samples varied across different pediatric cancers, thus indicating that disease biology may further play a role in altering the patients’ T cell developmental phenotype at collection44, which can inform collection practices as CAR T cells are applied to other diseases.
Another factor that can contribute to differences in T cell fitness lies in the previously exposed chemotherapy regimen. For example, chemotherapy regimens containing clofarabine or doxorubicin has been implicated in both quantitatively insufficient and poor-quality CAR T cell products 44,47 (we strongly discourage use of clofarabine prior to collection). Additionally, clinical data suggest that prior treatment with cyclophosphamide and cytarabine selectively reduces early lineage T cells that are associated with productive CAR T cell expansion 43. Therefore, it is important to understand how different chemotherapies affect T cells as it can have a direct impact on the quality of T cells collected. Early collection of T cells prior to intensive regimens of cytotoxic chemotherapy should be considered in patients identified as having a high risk of relapse or those with relapsed disease, which may improve the quality of the apheresis product and, thus, the resultant manufactured CAR T cell product.
Finally, differences in the CAR design and manufacturing processes may also play an important role in predicting the clinical performance of the final CAR T product. CAR T cell products that have shown efficacy in clinical trials to this point, including tisagenlecleucel, are second generation products 48-50. Tisagenlecleucel utilizes a 4-1BB based co-stimulatory domain and has been shown to persist in the blood for a median duration of 168 days (range 20-167 days) compared to CAR constructs using CD28 co-stimulatory domains, whose persistence is approximately 1 to 2 months 21,47,51. There are patients from the first CHOP studies with persistent CAR T cells for 5-10 years. This longer persistence is likely due to the reduced propensity for T cell exhaustion induced by tonic CAR signaling when co-stimulation is mediated by a 4-1BB domain 52. Data thus far suggests that CAR persistence is an important factor in achieving a durable remission in ALL without further anti-leukemia therapy. This association is harder to discern in lymphoma patients treated to date 53.