Non-thermal histotripsy tumor ablation promotes abscopal immune responses that enhance cancer immunotherapy

Abstract

Background: Developing the ability to use tumor-directed therapies to trigger potentially therapeutic immune responses against cancer antigens remains a high priority for cancer immunotherapy. We hypothesized that histotripsy, a novel non-invasive, non-thermal ablation modality that uses ultrasound-generated acoustic cavitation to disrupt tissues, could engender adaptive immune responses to tumor antigens.

Methods: Immunocompetent C57BL/6 mice inoculated with flank melanoma or hepatocellular carcinoma tumors were treated with histotripsy, thermal ablation, radiation therapy, or cytotoxic T lymphocyte-associated protein-4 (CTLA-4) blockade checkpoint inhibition. Lymphocyte responses were measured using flow cytometric and immunohistochemical analyses. The impact of histotripsy on abscopal immune responses was assessed in mice bearing bilateral tumors, or unilateral tumors with pulmonary tumors established via tail vein injection.

Results: Histotripsy ablation of subcutaneous murine melanoma tumors stimulated potent local intratumoral infiltration of innate and adaptive immune cell populations. The magnitude of this immunostimulation was stronger than that seen with tumor irradiation or thermal ablation. Histotripsy also promoted abscopal immune responses at untreated tumor sites and inhibited growth of pulmonary metastases. Histotripsy was capable of releasing tumor antigens with retained immunogenicity, and this immunostimulatory effect was associated with calreticulin translocation to the cellular membrane and local and systemic release of high mobility group box protein 1. Histotripsy ablation potentiated the efficacy of checkpoint inhibition immunotherapy in murine models of melanoma and hepatocellular carcinoma.

Conclusions: These preclinical observations suggest that non-invasive histotripsy ablation can be used to stimulate tumor-specific immune responses capable of magnifying the impact of checkpoint inhibition immunotherapy.

Keywords: immunology; oncology; tumors.

Figure 1

Histotripsy inhibition of tumor growth is well tolerated and promotes potent intratumoral CD8+ T cell infiltration. C57BL/6 mice were inoculated with B16GP33 subcutaneous tumors, then treated with sham (control) or histotripsy ablation (HT) on day 10. Subtotal histotripsy ablation resulted in durable suppression of tumor growth (A). No significant differences in body weight were observed following histotripsy ablation (B). Fluorescence-activated cell sorting (FACS) analysis of tumor-infiltrating lymphocytes (TIL) on day 3 (D3) and day 10 (D10) after treatment (gated on viable lymphocytes) demonstrated higher percentages of intratumoral CD8+ T cell infiltration after histotripsy by day 10 (C, D). Stimulation of CD8+ TIL remained significant on day 10 when cell numbers were normalized to tumor volume (E). FACS analysis of TIL on day 10 after treatment also identified marked increases in intratumoral populations of NK1.1+ natural killer cells, CD11c+ dendritic cells, Ly6G+ neutrophils, F4/80+ macrophages, and CD19+ B cells after histotripsy (F). Staining with major histocompatibility complex (MHC) tetramers loaded with the minor melanoma antigen GP33 and the major melanoma antigen GP100 identified significant increases in intratumoral populations of melanoma antigen-specific CD8+T cells (G). In contrast to histotripsy ablation (HT), tumor irradiation (XRT) and radiofrequency ablation (RFA) did not potentiate intratumoral CD8+ T cell infiltration (H) (n=4–8 per group; †p<0.05 vs all other groups by analysis of variance; *p<0.05 vs control by t-test).

Figure 2

Non-thermal histotripsy is capable of releasing immunogenic tumor neoantigens. C57BL/6 mice were inoculated with B16F10 or B16GP33 subcutaneous tumors, then treated with sham or histotripsy ablation (HT) on day 10. Tumors were excised and sham-treated tumors were liquefied using three freeze–thaw (F/T) cycles consisting of 2 min in liquid nitrogen and 2 min in a 60°C water bath. Liquefied tumors were exposed to CD8+ T cells isolated from the spleens of C57BL/6 mice 8 days after LCMV infection (at which time small populations of GP33-specific effector CD8+ T cells exist) in the presence of interleukin-2 and brefeldin A. Negative controls consisted of CD8+ T cells alone (no peptide) and positive controls consisted of CD8+ T cells exposed to 0.01 µg/mL GP33 peptide (GP33 peptide). After 5-hour incubation, individual (A) and group (B) fluorescence-activated cell sorting (FACS) analyses of intracellular IFNγ expression (gated on viable lymphocytes) demonstrated that exposure to histotripsy-treated B16GP33 tumors was the only experimental condition capable of stimulating GP33-specific CD8+ T cells in vitro. The magnitude of this stimulation was approximately half of that seen with direct exposure to GP33 peptide (n=4 per group, *p<0.05 vs negative control by t-test; †p<0.05 vs all other experimental groups by analysis of variance). IFNγ, interferon gamma; LCMV, lymphocytic choriomeningitis virus.

Figure 3

Histotripsy promotes regional and systemic tumor-specific CD8+ T cell responses. C57BL/6 mice were inoculated with B16GP33 subcutaneous tumors expressing very low levels of GP33 peptide, then treated with no therapy (control), tumor irradiation (XRT), radiofrequency ablation (RFA), or histotripsy ablation on day 10. 10 days later, fluorescence-activated cell sorting (FACS) analyses of splenocytes (gated on viable lymphocytes) demonstrated higher numbers of circulating CD8+ T cells with T cell receptor (TCR) specificity for the minor melanoma antigen GP33 after histotripsy but not after XRT or RFA (A, B). Histotripsy was also associated with a higher ratio of circulating melanoma-specific CD8+ T cells to regulatory CD4+ T cells (C) (n=4 per group, *p<0.05 vs control by t-test).

Figure 4

Histotripsy promotes abscopal CD8+ T cell responses. 10 days after bilateral B16GP33 melanoma tumor inoculation, C57BL/6 mice were treated with unilateral sham (control) or histotripsy tumor ablation. Fluorescence-activated cell sorting (FACS) analysis of tumor infiltrating lymphocyte (TIL) populations 10 days after treatment (gated on viable lymphocytes) identified stronger intratumoral CD8+ T cell infiltration in contralateral, untreated tumors after histotripsy (A). In contrast to histotripsy ablation (HT), unilateral tumor irradiation (XRT) and radiofrequency ablation (RFA) did not promote abscopal infiltration of CD8+ TIL (B). FACS analysis of histotripsy-ablated (HT ablated) and contralateral non-ablated tumors (HT abscopal) identified comparable levels of intratumoral CD8+ T cell infiltration (C). Immunohistochemistry (D) and immunofluorescence staining (E) of CD8 confirmed similar levels of CD8+ T cell infiltration between histotripsy-ablated (HT ablated) and contralateral non-ablated tumors (HT abscopal), but with different patterns of infiltration. Whereas CD8+ T cells were largely localized to peripheral areas that surrounded ablation zones in histotripsy-ablated tumors, CD8+ T cell infiltration was more diffuse and homogeneous in contralateral non-ablated tumors (D). Quantitation of CD8+ T cells using immunohistochemistry (defined as the average number of CD8+ T cells per 500 μm² area for a minimum of five non-overlapping fields from two independent samples) on days 3, 5 and 7 after sham or histotripsy ablation showed small and diminishing CD8+ T cell numbers over time in control tumors, rapidly increasing CD8+ T cell numbers in histotripsy-ablated tumors, and gradually increasing CD8+ T cell numbers in contralateral non-ablated tumors (F). Comparison of tumor growth between treated and contralateral non-treated tumors showed that histotripsy was associated with a modest but significant abscopal inhibition of contralateral tumor growth (G). In contrast, no abscopal effects were observed after unilateral irradiation (H) or radiofrequency ablation (I). This abscopal inhibition of contralateral tumor growth by histotripsy was associated with significant prolongation of survival (J) (n=4 per group and 23 per group in survival experiments, *p<0.05 vs control by t-test; †p<0.05 vs all other groups by analysis of variance).

Figure 5

Histotripsy inhibits the development of distant pulmonary metastases. C57BL/6 mice were inoculated with flank and intravenous inoculations of B16GP33 melanoma on days 0 and 3, respectively. Sham (control) or histotripsy (HT) of flank tumors was performed on day 10, and pulmonary metastases were quantified by visual inspection on day 20. Representative specimens (A) and group data (B) demonstrated significantly fewer pulmonary metastases in mice treated with flank tumor histotripsy ablation. Immunohistochemical analysis demonstrated smaller tumor size and more concentrated CD8+ T cell infiltration among pulmonary metastases in mice treated with histotripsy (C) (n=4 per group, *p<0.05 vs control by t-test; black bars indicate 50 µm).

Figure 6

Histotripsy promotes local and systemic inflammatory events and tumorous release of damage-associated molecular patterns. Ten days after B16GP33 melanoma tumor inoculation, C57BL/6 mice were treated with sham or histotripsy tumor ablation. Fluorescence-activated cell sorting (FACS) analysis of splenocytes (gated on viable lymphocytes) demonstrated significantly larger populations of circulating NK1.1+ natural killer cells and a trend toward larger populations of Ly6G+ neutrophils (A–C) 10 days after histotripsy. Immunofluorescence analyses of tumors confirmed colocalization of CRT with the endoplasmic reticulum protein ERp72 (data not shown); 1 day after histotripsy ablation, DAPI staining identified marked nuclear condensation and translocation of CRT to the plasma membrane (outlined in green) consistent with apoptosis induction (D). Immunofluorescence analysis of tumors also showed significant increases in levels of extracellular HMGB1 on day 1 (early) and day 7 (late) after histotripsy (treatment) (E). By day 7 after histotripsy, serum ELISA identified significantly higher levels of circulating HMGB1 (F) (n=3 per group, *p<0.05 vs control by t-test). CRT, calreticulin; HMGB1, high mobility group box protein 1.

Figure 7

Histotripsy augments the therapeutic efficacy of checkpoint inhibition. C57BL/6 mice bearing bilateral B16GP33 tumors were treated with or without anti-cytotoxic T lymphocyte-associated protein-4 (CTLA-4) mAb (CI) on days 6, 9 and 12, and with or without unilateral histotripsy tumor ablation (HT) on day 7. Serial tumor measurements demonstrated optimal suppression of non-ablated tumors in mice receiving combination therapy (A). When measured as a percentage of tumor infiltrating lymphocytes (TIL), intratumoral CD8+ T cells were increased in non-ablated tumors in all treatment groups (B); however, when measured as cell numbers normalized to tumor volume, significant increases in intratumoral CD8+ T cells within non-ablated tumors were only observed after combination therapy (C). C57BL/6 mice bearing bilateral Hepa1-6 tumors were treated with or without anti-CTLA-4 mAb on days 3, 6, 9 and 12, and with or without unilateral tumor ablation on day 10. Serial tumor measurements demonstrated optimal suppression of non-ablated tumors in mice receiving combination therapy (D) (n=4 per group; *p<0.05 vs control by t-test; †p<0.05 vs histotripsy by t-test; ‡p<0.05 vs all other groups by analysis of variance).