(P 119) EX VIVO EXPANDED TUMOR INFILTRATING LYMPHOCYTES FOR SARCOMA TREATMENT

Objective: Sarcomas are rare cancers that develop in the bone and soft tissues, with osteosarcoma being the most common primary malignant bone tumor, affecting both children and young adults. Despite advances in treatment, metastatic disease, particularly to the lungs, remains the most prominent cause of osteosarcoma-related deaths. Adoptive cell therapy using tumor-infiltrating lymphocytes (TILs) has shown durable clinical responses in metastatic melanoma patients and is being investigated for other solid tumors. However, TIL isolation and expansion in osteosarcoma and soft-tissue sarcomas is limited. Although TIL generation from sarcomas has been successful, the proportion of patients that can reliably generate a product of clinically meaningful number is lower (30%) than reported in melanoma (95%) along with inconsistently low tumor-specificities. Rapid expansion protocol (REP) the known method of expanding TILs has shown low success rate of TIL expansion (50%) from liposarcoma samples by our group. Discontinuation of anti-41BB antibody commercially along with high cost and variability of generating feeder cell layer for REP begs the need to develop a platform for expansion of TILs to clinical meaningful numbers. Artificial antigen presenting cells (AaPC) developed from K562 cells, expressing desired T-cell costimulatory molecules, are an appealing “off-the-shelf” feeder cells for human application. We have previously successfully utilized K562-AaPC to expand genetically modified T cells from peripheral blood for clinical trials. These K562 AaPC have also shown expansion of TILs comparable to REP from melanoma patients.

Methods: In the current study we used K562 AaPC expressing CD64, CD86, CD137L along with IL15/IL15Rα fusion protein (membrane-bound IL-15; mIL15) to expand TILs from sarcoma patients. Enzymatically dissociated single cell suspension from tumor tissue was co-cultured with irradiated OKT3 loaded K562 AaPC along with addition of exogenous cytokines (IL-2, and IL-21) in a 7-day stimulation cycle, and ex vivo expanded TILs were phenotypically and functionally characterized.

Results: Our data shows that K562 AaPCs have a 100% success rate of TIL expansion from sarcoma samples (n=5; Osteosarcoma=3, high grade undifferentiated sarcoma=1, sarcoma of left anterior shoulder=1) yielding (mean) 1.6e14 cells in (mean) 43.5 days of co-culture. K562 AaPC ex vivo expanded cells were predominantly CD8 (mean CD4/CD8, 0.09) and the platform generated a polyclonal TCR Vβ repertoire of the TIL product where all 24 Vβ families tested were preserved in the product. TILs generated were activated and capable of degranulation (mean CD107a, 35%) and anti-tumor response (mean IFN-g, 78.4%) upon stimulation. This was evident by targeting and lysis of autologous PDX tumors by TILs in a standard 4h chromium release cytotoxicity assay. TILs expressed memory cell markers (CCR7, CD62L, CD45RO) as well as determinants of an effector-cell phenotype. More than half of the TILs (mean, 88.7%) expressed CCR7, and half (mean, 48%) of all TILs expressed markers for central memory (Tcm, CD45RO+CD62L+). Co-expression of CCR7 and CD62L distinguishes a putative subset of memory T cells that have potential to home to lymph nodes.

Conclusion: These data support the hypothesis that K562 AaPCs are an ideal and superior ex vivo expansion platform for generation of activated TILs from sarcoma patients with memory phenotype which have implication for improved efficacy and long-term persistence after adoptive transfer.