Focal ablation technologies are routinely used in the clinical management of inoperable solid tumors but often result in incomplete ablations leading to high recurrence rates. Adjuvant therapies capable of safely eliminating residual tumor cells are therefore of great clinical interest. Interleukin 12 (IL-12) is a potent antitumor cytokine that can be localized intratumorally through coformulation with viscous biopolymers including chitosan (CS) solutions. The objective of this research was to determine if localized immunotherapy with CS/IL-12 could prevent tumor recurrence after cryoablation (CA). Tumor recurrence, overall survival, and protective immunity were assessed. Systemic immunity was evaluated in spontaneously metastatic and bilateral tumor models. Temporal bulk RNA sequencing was performed on tumor and draining lymph node samples. In multiple murine tumor models, the addition of CS/IL-12 to CA reduced recurrence rates by 30–55%. Altogether, this cryo-immunotherapy induced complete durable regression of large tumors in 80–100% of treated animals. Mice treated with CA plus adjuvant CS/IL-12 were partially or completely protected from tumor rechallenge. Systemically, CS/IL-12 prevented lung metastases when delivered as a neoadjuvant to CA. However, CA plus CS/IL-12 had minimal antitumor activity against established, untreated abscopal tumors. Adjuvant anti-PD-1 therapy delayed the growth of abscopal tumors. Transcriptome analyses revealed early immunological changes in the dLN, followed by a significant increase in gene expression associated with immune suppression and regulation. Cryo-immunotherapy with localized CS/IL-12 reduces recurrences and enhances the elimination of large primary tumors. This focal combination therapy also induces significant systemic antitumor immunity although further studies are necessary.

MC38 tumors were implanted subcutaneously in the right flanks of female C57BL/6 mice as described above. When tumor volumes measured between 200–500 mm³, CA was performed and CS/IL-12 was intratumorally injected within an hour after CA. Mice were euthanized 3 days and 7 days post-ablation to harvest tumor tissues and tumor-draining inguinal lymph nodes (dLNs). To obtain cell suspensions, dLNs were homogenized through 40 µm mesh strainers with complete T cell media. Tumor tissues were first minced into approximately 0.5 cm squares before treatment with 5X Triple Enzyme Solution (collagenase type IV 10 mg/ml, hyaluronidase 1 mg/ml, and DNase 20,000 units final concentration in Hank’s Buffered Saline Solution) for 30 minutes at 37°C with magnetic bar stirring. The resulting homogenate was filtered through a 70 µm mesh strainer into a clean tube to achieve the final cell suspension.

Total RNA was extracted from the cell suspension using the GeneJET RNA Purification Kit (ThermoFisher cat #K0731) according to the manufacturer’s protocol. Briefly, Lysis Buffer supplemented with β-mercaptoethanol and then ethanol (99%) was added to the cell suspension and vortexed. The lysate was then loaded onto the purification spin column and washed twice before eluting with nuclease-free water. Samples were sent to BGI Genomics (Hong Kong, China) to perform transcriptome sequencing using their DNBSEQ^TM platform using a read length of 100 paired-end base pairs. Data analysis was performed using the BGI software platform, Dr. Tom.

Sample inclusion was based on RNA quality (260/280 > 1.8) and successful library construction. Only one sample, from a tumor treated with CA + CS/IL-12 (11_T_D3), failed to meet these criteria and was omitted from further analysis. The dLN from one mouse treated with CA alone and then euthanized on day 7 after treatment was not recovered during necropsy (34_dLN_D7). Secondary exclusion criteria were based on read alignment coverage and randomness. The technical difficulty of extracting high-quality RNA from highly necrotic ablated tumor samples, in particular three days after ablation, hindered the interpretation of gene expression from the tumor at day 3 post-treatment.

Unless otherwise noted, for all gene expression analysis, genes were filtered from the transcriptome using a gate of Q < 0.05 for the comparisons of interest. Bowtie2 was used to map the clean reads to the reference gene sequence (transcriptome), and then RSEM was used to calculate the gene expression level of each sample. Differential gene detection was calculated using the DESeq2 method, where Q is calculated using the FDR/Benjamini-Hochberg method. According to the results of differential gene detection, the R package pheatmap was used to perform hierarchical clustering analysis on the union set differential genes.

Excel file format, or .csv reader