Tumor infiltrating lymphocyte therapy for renal cell carcinoma: methods for ex vivo expansion and T-cell memory induction

Abstract

Tumor infiltrating lymphocyte (TIL) therapy is a personalized treatment for locally advanced or metastatic cancer. TIL therapy requires the surgical excision of a portion of the malignant tissue and subsequent ex vivo expansion of tumor reactive immune cells, primarily T-lymphocytes (T-cells), prior to reinfusion into the patient. TIL therapy has been shown to be efficacious in numerous forms of cancer; however, a lack of consistent in vitro TIL production has hindered the development of this potentially curative cell therapy for renal cell carcinoma (RCC). I (Mitchell W. Braun) report my work, in the Godwin research laboratory, which has aimed to advance the prospect of TIL therapy for RCC. A novel method for TIL expansion has been developed that has shown a 94% success rate for creating TIL cultures from clear cell RCC. This method generates a TIL product with an optimal phenotype relative to other established TIL production protocols, namely the pre-rapid expansion protocol (PreREP)/REP and FTD + beads method. PreREP involves IL-2 supplementation alone as the source of exogenous stimulation during the first phase of expansion. The fresh tumor digests (FTD) + beads method involves supplementation of IL-2, as well as anti-CD3/CD28 beads to initiate TIL cultures. Relative to the PreREP and FTD + beads protocols, our method involves adherent cell depletion (ACD), referred to as panning, and generates significantly fewer regulatory (CD4+/CD25+/FOXP3+) (p=0.049, p=0.005), tissue-resident memory (CD8+/CD103+) (p=0.027, p=0.009), PD1+/TIM-3+ double-positive (p=0.009, p=0.011) and TIGIT+ T-cells (p=0.049, p=0.026), respectively. These phenotypic changes were achieved while increasing the average tumor reactive T-cell yield in the final TIL product. However, using this method a majority of the CD8+ expanded TILs bear an effector-memory or effector phenotype, which is sub-iv optimal for adoptive transfer and similar to the other protocols. Therefore, various methods to induce a less differentiated phenotype during in vitro TIL expansion were. AKT inhibition (AKTi) during TIL expansion was shown to increase expression of the memory-associated/lymph-node homing selectin, CD62L (p<0.0001), which would be expected to increase TIL effectiveness and the durability of responses. However, AKTi decreased TIL yield after two weeks of expansion (p=0.047). To identify a pathway that might mimic the advantageous effects of AKTi but avoid the loss of TILs, a reverse phase protein array (RPPA) was used to identify downstream targets of AKTi. This analysis identified the kinase GSK3β as a major down-regulated target of AKTi. Inhibition of GSK3β is known to upregulate the pioneer transcription factor TCF1, which is associated with T-cell memory development. Subsequent investigation demonstrated that treatment with a GSK3β inhibitor (GSK3βi) during TIL expansion did indeed recapitulate the positive effect of increasing expression of CD62L (p=0.029). Importantly, though, GSK3βi did not affect TIL yield (p=0.76). This approach has the potential to increase the persistence of the TIL product after infusion leading to increased durable response rates in RCC and potentially other tumors. I submit these findings, as well as additional data, as part of my application for the degree of Doctor of Philosophy in Pathology and Laboratory Medicine from the University of Kansas Medical Center.