Comparative Analysis of the Immune Response to Hot and Cold Ablation in a Colon Cancer Mouse Model

Objective: To characterize and compare radiofrequency ablation (RFA) and cryo-ablation (CRA) immune modulation in a colon cancer mouse model.

Methods: All studies were conducted under a research protocol approved by the Institutional Animal Care and Use Committee. Balb/c mice were inoculated with CT26 cells, a syngeneic mouse colon carcinoma line. Mice were randomized to be treated with RFA (N=10, 22G, Baylis), CRA (N=10, 18.5G, Galil, BTG), or sham (N=6) when tumors reached a volume of at least 50 mm3. Serum samples were taken 6 hours after the ablation procedure via retro-orbital bleed. Mice were sacrificed 3 days post-treatment, and the tumor, spleen, and serum were harvested. Tumors were embedded in paraffin and sliced for immunohistochemical analysis of cell death and apoptosis (Caspase3), thermal damage (HSP70), apoptotic DNA fragmentation (TUNEL), and immune cell infiltration (CD3, CD8, NK, T-reg). Splenocyte cell suspensions were stained with multiple antibodies and expression of various sub-populations of immune cells were measured via multiparameter flow cytometry (BD FACSymphony). Circulating cytokines related to pro- and anti-inflammatory responses were measured (V-PLEX Panel). Differences between each treatment and sham were calculated using two tailed t-tests.

Results: At 3 days post ablation, both RFA and CRA increased Caspase3 expression 6-fold compared to sham (p<0.005) and reduced HSP70 expression 10-fold in the tumor microenvironment (TME) (p<0.06). RFA alone induced DNA fragments as measured with TUNEL staining 2.3-fold (p<0.01). Immune marker stains in the TME revealed no differences in T cell (CD3) or T-cytotoxic (CD8) expression between treatment groups and sham. However, both RFA and CRA had 3.8-fold more natural killer cells (NK) compared to sham (p=0.08), and RFA alone reduced T-regulatory cells 6-fold, (p<0.0001) in the TME. Flow cytometry revealed 1.75-fold more CD8 in the RFA and CRA cohorts compared to sham (p<0.05) but only RFA increased CD4 (T-Helper cells) with a 1.55-fold increase(p<0.05).
Cytokine analysis at 6 hours post CRA treatment demonstrated 2.7-, 3.5-, and 15.3-fold higher levels of the pro-inflammatory cytokines IL-1B, KC GRO, and IL6, respectively (p<0.05). The CRA and RFA cohorts showed 1.5- and 1.7-fold higher immune suppressor cytokine IL-10 levels 6 hours post- treatment, respectively (p<0.05).

Conclusions: Three components of the immune system were analyzed: tumor microenvironment (TME), systemic response, and cytokine release. Interestingly CRA augmented a wider range of cytokines than RFA, which were mainly pro-inflammatory but also anti-inflammatory. In the TME component, RFA induced DNA fragmentation and demonstrated a slightly more effective immune augmentation as reflected by increased NK and decreased T-reg expression. There was an increase in expression of CD4 and CD8 cells in the spleen that was not observed in the TME of treated tumors which may suggest that systemic immune analysis might not reflect the immune infiltrates in the TME. Complete and comprehensive investigation of immune response to locoregional therapy should ideally include multiple immune compartments, components, and timepoints. A broad temporal and spatial sampling may improve our understanding of immune mechanisms, and in turn better inform the development of novel drugs and drug-device combinations to optimize potentiation of immune response.