J. R. De La Torre Medina1,2, J. Shankara Narayanan2, S. Abeynaike3, S. Paust3, R. White1,2 1University Of California – San Diego, General Surgery, San Diego, CA, USA 2Moores Cancer Center, Surgical Oncology, La Jolla, CA, USA 3The Scripps Research Institute, Immunology, La Jolla, CA, USA
Introduction:
Immunotherapy has shown huge benefits in other solid tumors, but results in pancreatic cancer (PC) have been disappointing. We have shown that irreversible electroporation (IRE), a non-thermal form of tumor ablation, can induce systemic immune responses in murine PC models. Patient-derived xenograft (PDX) models capture the heterogeneity of human PC but require transplantation into immunodeficient mice. We hypothesize that transplanted human tumor-infiltrating lymphocytes (TILs) can transiently engraft and permit assessment of immune responses to IRE in PDX models.
Methods:
A freshly-resected PC tumor from a female subject status post neoadjuvant chemotherapy was minced, and 2 mm3 undisrupted pieces of tumor were implanted into SQ pockets of gender-matched immunodeficient NOD scid gamma (NSG) mice (n = 20) without accompanying transfer of any other patient-derived cells. Palpable tumors were observed in 7/20 mice within 12 weeks of implantation. Six mice were randomized equally between control and IRE treatment groups (150 x 500 microsecond pulses at 1500 V/cm). The tumors were harvested and analyzed by flow cytometry at 14 days post-procedure.
Results:
The average tumor volume was 68±32 mm3 before randomization which was reduced to 41±22 mm3 two weeks post-IRE (including one complete response) and grew to a volume of 550±109 mm3 in untreated mice. There was a statistically significant difference in post-euthanasia tumor volumes and weights between the IRE and control groups (Table). Flow cytometry showed a robust infiltration of human CD45+ (hematopoietic) cells within all 5 tumors available for analysis, demonstrating persistence of human immune cells. Although control tumors had more CD45+ cells of human origin than the IRE group (Table), IRE-treated tumors had more cytotoxic (CD8+) T-cells and more activated (CD8+) T-cells that express interferon-gamma (IFNg) than the control group. The differences in TILs by flow cytometry did not reach statistical significance due to the small sample sizes.
Conclusion:
This is the first demonstration that IRE is effective at controlling tumor growth in a PDX model. More importantly, human TILs persist in transplanted tumors, and IRE is capable of increasing the proportion of cytotoxic and activated cytotoxic T-cells within the tumors. PDX models can therefore be used to evaluate the effects of IRE alone and in combination with immunotherapy across the whole spectrum of human PC.
