Targeting STAT3 to Suppress Systemic Pro-Oncogenic Effects from Hepatic Radiofrequency Ablation

Abstract

Purpose To (a) identify key expressed genes in the periablational rim after radiofrequency ablation (RFA) and their role in driving the stimulation of distant tumor growth and (b) use adjuvant drug therapies to block key identified mediator(s) to suppress off-target tumorigenic effects of hepatic RFA. Materials and Methods This institutional animal care and use committee-approved study was performed in C57BL6 mice (n = 20) and F344 rats (n = 124). First, gene expression analysis was performed in mice after hepatic RFA or sham procedure; mice were sacrificed 24 hours to 7 days after treatment. Data were analyzed for differentially expressed genes (greater than twofold change) and their functional annotations. Next, animals were allocated to hepatic RFA or sham treatment with or without STAT3 (signal transducer and activator of transcription 3) inhibitor S3I-201 for periablational phosphorylated STAT3 immunohistochemistry analysis at 24 hours. Finally, animals with subcutaneous R3230 adenocarcinoma tumors were allocated to RFA or sham treatment with or without a STAT3 inhibitor (S3I-201 or micellar curcumin, eight arms). Outcomes included distant tumor growth, proliferation (Ki-67 percentage), and microvascular density. Results At 24 hours, 217 genes had altered expression (107 upregulated and 110 downregulated), decreasing to 55 genes (27 upregulated and 28 downregulated) and 18 genes (four upregulated, 14 downregulated) at 72 hours and 7 days, respectively. At 24 hours, STAT3 occurred in four of seven activated pathways associated with pro-oncogenic genes at network analysis. Immunohistochemistry analysis confirmed elevated periablational phosphorylated STAT3 24 hours after RFA, which was suppressed with S3I-201 (percentage of positive cells per field: 31.7% ± 3.4 vs 3.8% ± 1.7; P < .001). Combined RFA plus S3I-201 reduced systemic distant tumor growth at 7 days (end diameter: 11.8 mm ± 0.5 with RFA plus S3I-201, 19.8 mm ± 0.7 with RFA alone, and 15 mm ± 0.7 with sham procedure; P < .001). STAT3 inhibition with micellar curcumin also suppressed postablation stimulation of distant tumor growth, proliferation, and microvascular density (P < .01). Conclusion Gene expression analysis identified multiple pathways upregulated in the periablational rim after hepatic RFA, of which STAT3 was active in four of seven. Postablation STAT3 activation is linked to increased distant tumor stimulation and can be suppressed with adjuvant STAT3 inhibitors. ^© RSNA, 2017.

Figure 1:

Altered expression of 217 genes in periablational rim 24 hours after hepatic RFA. Heat map delineates changes up- and downregulated DEGs between three samples each from periablational rim 24 hours after hepatic RFA and with sham treatment (fold change ≥1.5, false discovery rate–adjusted P < .05). Rows and columns are sorted on the basis of hierarchical clustering. Dendogram (top, x-axis) was based on hepatic RFA and sham samples at 24 hours and on correlation between gene expression of difference (left, y-axis). Color bar (top left) demonstrates log2-fold changes from comparison between periablational rim at 24 hours after RFA versus sham treatment. Green to red colors represent log2-fold change from low expression to high expression (+2 to −2 FC). Upregulation is displayed in shades of red, downregulation in green, and mean gene expression in black.

Figure 2:

Network analysis for STAT3 related to upregulated genes in periablational rim 24 hours after hepatic RFA. Network map of STAT3 gene demonstrates its association with various genes upregulated after hepatic RFA, including those with known roles in tumor growth. Genes activated in our analysis are outlined in black. Genes linked to postablation tissue reactions, though not upregulated in our analysis, are outlined in red.

Figure 3a:

Increased periablational phosphorylated STAT3 expression after hepatic RFA. Images from immunohistochemistry analysis obtained 24 hours after RFA and sham procedure demonstrate significantly increased periablational phosphorylated STAT3 (arrow in b and d) after RFA in both C57Bl6 mice (a, b, model used for gene expression studies) and F344 rats (c, d). No measureable phosphorylated STAT3 staining was observed.

Figure 3b:

Increased periablational phosphorylated STAT3 expression after hepatic RFA. Images from immunohistochemistry analysis obtained 24 hours after RFA and sham procedure demonstrate significantly increased periablational phosphorylated STAT3 (arrow in b and d) after RFA in both C57Bl6 mice (a, b, model used for gene expression studies) and F344 rats (c, d). No measureable phosphorylated STAT3 staining was observed.

Figure 3c:

Increased periablational phosphorylated STAT3 expression after hepatic RFA. Images from immunohistochemistry analysis obtained 24 hours after RFA and sham procedure demonstrate significantly increased periablational phosphorylated STAT3 (arrow in b and d) after RFA in both C57Bl6 mice (a, b, model used for gene expression studies) and F344 rats (c, d). No measureable phosphorylated STAT3 staining was observed.

Figure 3d:

Increased periablational phosphorylated STAT3 expression after hepatic RFA. Images from immunohistochemistry analysis obtained 24 hours after RFA and sham procedure demonstrate significantly increased periablational phosphorylated STAT3 (arrow in b and d) after RFA in both C57Bl6 mice (a, b, model used for gene expression studies) and F344 rats (c, d). No measureable phosphorylated STAT3 staining was observed.

Figure 4a:

Charts show periablational and distant tumor VEGF and HGF levels with sham procedure, RFA without and with adjuvant STAT3 inhibitor (SI), and STAT3 inhibitor alone. (a, b) Hepatic RFA increased HGF levels in periablational rim (a) and distant tumor (b). Adjuvant STAT3 inhibition with S3I-201 did not suppress RFA-induced changes in local or distant HGF levels. Concentration is in picograms per milliliter. (c, d) Conversely, hepatic RFA increased periablational rim (c) and distant tumor (d) VEGF levels, which were suppressed with adjuvant STAT3 inhibition—suggesting that RFA-induced increases in VEGF are mediated through STAT3.

Figure 4b:

Charts show periablational and distant tumor VEGF and HGF levels with sham procedure, RFA without and with adjuvant STAT3 inhibitor (SI), and STAT3 inhibitor alone. (a, b) Hepatic RFA increased HGF levels in periablational rim (a) and distant tumor (b). Adjuvant STAT3 inhibition with S3I-201 did not suppress RFA-induced changes in local or distant HGF levels. Concentration is in picograms per milliliter. (c, d) Conversely, hepatic RFA increased periablational rim (c) and distant tumor (d) VEGF levels, which were suppressed with adjuvant STAT3 inhibition—suggesting that RFA-induced increases in VEGF are mediated through STAT3.

Figure 4c:

Charts show periablational and distant tumor VEGF and HGF levels with sham procedure, RFA without and with adjuvant STAT3 inhibitor (SI), and STAT3 inhibitor alone. (a, b) Hepatic RFA increased HGF levels in periablational rim (a) and distant tumor (b). Adjuvant STAT3 inhibition with S3I-201 did not suppress RFA-induced changes in local or distant HGF levels. Concentration is in picograms per milliliter. (c, d) Conversely, hepatic RFA increased periablational rim (c) and distant tumor (d) VEGF levels, which were suppressed with adjuvant STAT3 inhibition—suggesting that RFA-induced increases in VEGF are mediated through STAT3.

Figure 4d:

Charts show periablational and distant tumor VEGF and HGF levels with sham procedure, RFA without and with adjuvant STAT3 inhibitor (SI), and STAT3 inhibitor alone. (a, b) Hepatic RFA increased HGF levels in periablational rim (a) and distant tumor (b). Adjuvant STAT3 inhibition with S3I-201 did not suppress RFA-induced changes in local or distant HGF levels. Concentration is in picograms per milliliter. (c, d) Conversely, hepatic RFA increased periablational rim (c) and distant tumor (d) VEGF levels, which were suppressed with adjuvant STAT3 inhibition—suggesting that RFA-induced increases in VEGF are mediated through STAT3.

Figure 5a:

Adjuvant S3I-201, a STAT3 inhibitor (SI), blocks early (0–7 days) hepatic RFA-induced distant tumor growth and increases overall animal end point survival. R = RFA alone, S = sham treatment. (a) Graph shows that adjuvant S3I-201 given immediately after hepatic RFA reduces distant R3230 tumor growth compared with hepatic RFA alone (R) and to levels even lower than that in animals that received sham procedure (P < .01). The S3I-201 alone treatment arm has same growth rate as sham arm, indicating that STAT3 inhibitor does not have an active effect on tumor growth by itself. (b) In long-term animal end point study (implanted subcutaneous tumor size, 20 mm), the survival time of animals treated with hepatic RFA (mean, 13.0 days ± 0.0) was shorter than that with sham treatment (17.0 days ± 0.8) (P = .04). A single dose of STAT3 inhibitor (SI-OH) with RFA increased survival to 19.8 days ± 0.5 (P < .04 vs RFA and sham treatment), whereas RFA with multiple doses of S3I-201 (every 3 days, E3D ) showed greatest mean survival (21.5 days ± 1.0, P < .04 for all comparisons).

Figure 5b:

Adjuvant S3I-201, a STAT3 inhibitor (SI), blocks early (0–7 days) hepatic RFA-induced distant tumor growth and increases overall animal end point survival. R = RFA alone, S = sham treatment. (a) Graph shows that adjuvant S3I-201 given immediately after hepatic RFA reduces distant R3230 tumor growth compared with hepatic RFA alone (R) and to levels even lower than that in animals that received sham procedure (P < .01). The S3I-201 alone treatment arm has same growth rate as sham arm, indicating that STAT3 inhibitor does not have an active effect on tumor growth by itself. (b) In long-term animal end point study (implanted subcutaneous tumor size, 20 mm), the survival time of animals treated with hepatic RFA (mean, 13.0 days ± 0.0) was shorter than that with sham treatment (17.0 days ± 0.8) (P = .04). A single dose of STAT3 inhibitor (SI-OH) with RFA increased survival to 19.8 days ± 0.5 (P < .04 vs RFA and sham treatment), whereas RFA with multiple doses of S3I-201 (every 3 days, E3D ) showed greatest mean survival (21.5 days ± 1.0, P < .04 for all comparisons).