WO2020089209A1 - Treatment of cancer using kinase inhibitors - Google Patents

Abstract

The application relates to methods of treating cancer in a patient using kinase inhibitors, as well as methods of predicting the response of a cancer to treatment with a kinase inhibitor. In one embodiment, the method comprises determining the level of macrophages, and optionally the level of NK cells, in a tumour sample obtained from the patient. In addition, or alternatively, a method of predicting the response of a cancer to treatment with a kinase inhibitor may comprise determining the level of interleukin-1 β (IL-Ιβ) and/or interleukin-18 (IL-18) in a sample obtained from the patient.

Description

Treatment of Cancer Using Kinase Inhibitors

Field

The present invention relates to methods of treating cancer in a patient using kinase inhibitors, as well as methods of predicting the response of a cancer to treatment with a kinase inhibitor. In one embodiment, the method comprises determining the level of macrophages, and optionally the level of NK cells, in a tumour sample obtained from the patient. In addition, or alternatively, a method of predicting the response of a cancer to treatment with a kinase inhibitor may comprise determining the level of interleukin-1 b (I L-1 b) and/or interleukin-18 (I L-18) in a sample obtained from the patient.

Liver cancer is the second leading cause of cancer-related deaths worldwide (Ferlay et ai, 2015). Hepatocellular carcinoma (HCC) is the most common liver cancer subtype with increasing incidence and poor prognosis. Hepatitis B and C viral infection, alcoholism as well as non-alcoholic steatohepatitis are the predominant risk factors for HCC development (Janevska et al., 2015).

Sorafenib, a broad spectrum kinase inhibitor, was approved for the treatment of patients with unresectable HCC in 2007 (Keating et ai, 2017). The small molecule targets BRAF, CRAF, MAP kinases, VEGFR and PDGFR, resulting in apoptosis of tumour cells and inhibition of angiogenesis (Wilhelm et ai, 2008). Sorafenib has been shown to prolong the survival of HCC patients by 3 months compared to placebo-treated patients (Llovet et a , 2008).

However, the efficacy of sorafenib treatment across HCC patient populations varies and durable, as well as complete responses, have been described only for a minority of advanced HCC patients (Yada et ai, 2014). In addition, sorafenib treatment causes side effects such as hypertension, diarrhoea or hand-foot skin reactions (Keating et ai, 2017).

In view of the low response rate and severe side effects associated with sorafenib treatment in HCC patients, methods for identifying patients likely to benefit from sorafenib treatment have been proposed. One method employed relies on the detection of early dermatologic reactions which have been shown to predict better survival of patients receiving sorafenib (Rimola et ai, 2017; Reig et ai, 2014; Branco et ai, 2017). In addition, a method based on the detection of higher serum cytokine levels has been proposed in order to identify patients predicted to benefit from sorafenib treatment (Hayashi et al., 2017). The detection of side effects associated with sorafenib, specifically hand-foot syndrome and diarrhoea, has also been suggested to predict the clinical efficacy of sorafenib in HCC patients (Cho et al., 2013). However, all of these methods have the downside that they can only be employed after treatment with sorafenib has commenced.

Thus, there remains a need in the art for methods for identifying patients which are likely to respond to treatment with kinase inhibitors, such as sorafenib, in particular methods which can be used to identify patients likely to benefit from treatment before treatment has commenced.

Sorafenib, a type II kinase inhibitor, is known to mediate its effect against HCC via antiangiogenic and cytotoxic effects. Indications of additional immunomodulatory effects mediated by sorafenib also existed but these were not fully understood. These

immunomodulatory effects have now been elucidated by the present inventors who have shown that sorafenib induces pyroptosis of tumour-associated macrophages (TAMs) present in HCC tumours, leading to cytokine release, cytotoxic NK cell activation, tumour infiltration and tumour cell killing. Similar effects were also observed with the type I kinase inhibitor sunitinib.

TAMs are known to be present in the microenvironment of solid tumours. In view of their role in tumour progression, high TAM infiltration levels have been shown to be indicators of poor prognosis in a number of cancers, including lung cancer, oesophageal squamous cell carcinoma and gastric cancer. Therapeutic strategies to reduce the presence of TAMs in the tumour microenvironment have also been proposed, including depleting TAMs, reducing TAM recruitment into the tumour tissue, and reprogramming TAMs into anti-tumour macrophages.

Based on the finding by the present inventors that kinase inhibitors such as sorafenib and sunitinib induce pyroptosis of TAMs present in the tumour microenvironment, it is expected that tumours comprising high TAM infiltration levels are more likely to respond to kinase inhibitor treatment than tumours comprising low levels of TAMs. In addition, tumours comprising high levels of NK cells are similarly expected to be more likely to respond to sorafenib treatment than tumours comprising low levels of NK cells. However, as the present inventors have also shown that NK cells are also recruited into the tumour as a result of macrophage pyroptosis, NK cell levels are expected to be a secondary indicator for identifying tumours likely to respond to kinase inhibitor treatment.

As explained above, the detection of early dermatologic reactions and higher serum cytokine levels has been proposed to identify patients likely to benefit from sorafenib treatment but these tests have the disadvantage that they can only be employed after the start of sorafenib treatment. In contrast, methods involving determining the level of TAMs, and optionally NK cells, present in a tumour to predict the likely response of the tumour to kinase inhibitors such as sorafenib and sunitinib can be employed before treatment has commenced, thereby avoiding treatments which would ultimately prove not effective for the patient, as well as the side effects associated therewith.

Thus, in a first aspect the present invention relates to a kinase inhibitor for use in a method of treating cancer in a patient, wherein a tumour of the patient has been determined to comprise a level of macrophages above a predetermined threshold. In a preferred embodiment, the method comprises determining the level of macrophages in a tumour sample obtained from the patient and comparing the level of macrophages to the

predetermined threshold.

Method of predicting the response of a cancer to treatment with a kinase inhibitor are also contemplated. The method preferably comprises determining the level of macrophages in a tumour sample obtained from the patient and comparing the level of macrophages to a predetermined threshold, wherein a level of macrophages in the tumour sample which exceeds the threshold indicates that the patient will respond to treatment with the kinase inhibitor.

The macrophage levels detected are preferably, levels of tumour associated macrophages (TAMs). TAMs have been shown to predominantly comprise macrophages of the M2 subtype, which are known to promote tumour development. The present inventors have demonstrated that sorafenib is capable of inducing pyroptosis of macrophages regardless of the macrophage subtype. Thus, determination of TAM levels may comprise determination of the level of M2, M1 , and M2c TAMs.

The predetermined threshold is a threshold which is suitable for identifying tumours that are likely to respond to treatment with a kinase inhibitor. As explained above, it is expected that tumours comprising higher levels of macrophages are more likely to respond to kinase inhibitor treatment that tumours comprising lower levels of macrophages. The predetermined threshold with which the macrophage level in a tumour is compared may be the 75^th percentile of the distribution of macrophage levels in tumours of a cancer of interest, such as HCC. In this case, a tumour comprising a macrophage level above the 75^th percentile indicates that the tumour will respond to treatment with the kinase inhibitor.

As noted above, NK cell levels present in a tumour are also expected to be predictive of the response of the tumour to treatment with a kinase inhibitor. Thus, a method as described herein may further comprise determining the level of NK cells in a tumour sample obtained from a patient and comparing the level to a predetermined threshold. The predetermined threshold is a threshold which is suitable for identifying tumours that are likely to respond to treatment with a kinase inhibitor. The predetermined threshold with which the NK cell level in a tumour is compared may be the 75^th percentile of the distribution of NK cell levels in tumours of a cancer of interest, such as HCC. In this case, a tumour comprising an NK cell level above the 75^th percentile indicates that the tumour will respond to treatment with the kinase inhibitor.

The macrophage level may be determined in a tumour sample obtained from a patient. In context of methods comprising determination of the macrophage and/or NK cell level in a tumour, the tumour sample is preferably a tumour sample obtained from a patient prior to the administration of the kinase inhibitor to the patient. Thus, the patient is preferably a patient which has not previously been treated with the kinase inhibitor.

The kinase inhibitor is a kinase inhibitor capable of inducing pyroptosis of macrophages, in particular TAMs, upon administration of the kinase inhibitor to a cancer patient. In a preferred embodiment, the kinase inhibitor is a type II kinase inhibitor such as sorafenib, or a type I kinase inhibitor such as sunitinib, or a pharmaceutically acceptable salt, hydrate, or solvate thereof. Most preferably, the kinase inhibitor is sorafenib, or a pharmaceutically acceptable salt, hydrate, or solvate thereof.

The cancer may be any cancer responsive to treatment with a kinase inhibitor, such as sorafenib and sunitinib. The cancer is preferably a solid cancer as only solid cancers comprise TAMs. Sorafenib has been shown to be effective in the treatment of HCC, renal cell carcinoma (RCC), or differentiated thyroid cancer (DTC). Thus, in a preferred embodiment, in particular where the kinase inhibitor is sorafenib, the cancer may be selected from the group consisting of HCC, RCC and DTC but preferably is HCC. Sunitinib has been shown to be effective in the treatment of gastrointestinal stromal tumor (GIST), advanced RCC, and advanced pancreatic neuroendocrine cancer. Thus, where the kinase inhibitor is sunitinib, the cancer may be selected from the group consisting of: GIST, RCC and neuroendocrine cancer of the pancreas.

Pyroptosis is known to be associated with cytokine release, including the release of IL-1 b and IL-18 (Bergsbaken et al., 2009). The present inventors similarly detected the release of I L-1 b and IL-18 as a result of macrophage pyroptosis following sorafenib treatment. Thus, it is expected that high levels of I L-1 b and IL-18 in a cancer patient following treatment with a kinase inhibitor, such as sorafenib or sunitinib, is indicative of higher levels of macrophage pyroptosis and thus a higher likelihood that the tumour will respond to the treatment.

Thus, in a second aspect the present invention provides a method of predicting the response of a cancer patient to treatment with a kinase inhibitor, wherein the method comprises: administering the kinase inhibitor to the patient;

determining the level of IL-1 b and/or IL-18 in a sample obtained from the patient; and comparing the level of I L-1 b and/or IL-18 to a predetermined threshold,

wherein a level of I L-1 b and/or IL-18 which exceeds the threshold indicates that the patient will respond to treatment with the kinase inhibitor.

For example, the method may comprise determining the level of I L-1 b and IL-18, whereby the level of I L-1 b and the level of IL-18 are each compared to a predetermined threshold.

The predetermined threshold is preferably the baseline level of I L-1 b and IL-18 in the patient prior to first administration of the kinase inhibitor to the patient.

The sample in which the level of IL-1 b and/or IL-18 is determined in a sample, such as a serum sample, obtained from the patient after the administration of the kinase inhibitor to the patient. The detection of cytokines levels in e.g. serum samples, has the advantage that the method is less invasive than methods requiring tumour samples.

In another aspect the invention relates to a method combining a method according to the first aspect and a method according to the second aspect, whereby the method according to the first aspect is carried out prior to administration of the kinase inhibitor to the patient and the method according to the second aspect is carried out after administration of the kinase inhibitor to the patient. Such a method may identify patients likely to benefit from kinase inhibitor treatment with greater accuracy than the performance of either method alone. Brief Description of Figures

Figure 1 : Sorafenib treatment of iAST mice leads to strong tumour growth inhibition.

(a) Treatment schedule: Sorafenib therapy started at day 54 after virus injection and was continued for 8 days (90 mg/kg, p.o., daily). Necropsy of vehicle (Co) and sorafenib (Sora) groups (n = 5) was performed on day 9 after treatment initiation (b) Sorafenib-treated mice show a significant lower total liver weight (liver incl. multinodular tumours) compared to control mice at day 9. (c) The IHC quantification of Ki67⁺ cells of tumour sections show no differences in the proliferation between vehicle- and sorafenib-treated tumours (n = 4). In the graph, each dot represents the analysis of multiple tumours in 1 liver section of a mouse (d) The quantification of the caspase-3⁺ area in tumour sections indicates an increase of apoptotic cells after sorafenib treatment (n = 5). In the graph, each dot represents the average of 4 images per tumour (b, c, d) Comparisons between groups were performed by Student’s t-test (^***p<0.001 ).

Figure 2: Sorafenib therapy modulates intratumoural innate immune cell populations and induces pro-inflammatory cytokine release. iAST mice were treated with vehicle (Co) or 90 mg/kg sorafenib (Sora) p.o. daily as shown in Figure 1 a. The immune infiltrate in the tumours was analyzed via flow cytometry at day 9 of treatment (a) Macrophage (MF) density decreases after sorafenib therapy (n = 5). Phenotypic analysis indicates a dominant population of the M1 -type MF within the F4/80⁺ population independent of therapy (b) The IHC quantification of the F4/80⁺ area in tumour sections confirms the reduction of intratumoural MF after sorafenib treatment (n = 4). Each dot represents the analysis of multiple tumours in 1 liver section of a mouse. Representative IHC images of F4/80⁺ staining for each group (c) NK cells increase with sorafenib treatment and exhibit upregulated markers of activation tumours (CD69 and PD-1 ), proliferation (Ki67) and degranulation (CD107a). (d) T cell infiltration (CD4⁺ and CD8⁺) shows no difference between control and sorafenib treatment (e) Quantification of CD3⁺ T cells in immunofluorescence (IF) sections. Each dot represents the average of 5 images per tumour. Representative IF images confirm no change in T cells within control and sorafenib tumours (f) Analysis of the total immune infiltrate (CD45⁺), DC and MDSC show no significant changes between control and sorafenib cohorts (g) Cytokine analysis of tumour lysates show a significant upregulation of pro- inflammatory cytokine secretion after 16 hours of treatment (a - g) Comparisons between groups were performed by Student’s t-test (^*p<0.05, ^**p<0.01 , ^***p<0.001 ).

Figure 3: Depletion of macrophages abolishes the therapeutic effect of sorafenib in iAST tumours, (a) Treatment schedule indicating the days of cell depletion (MF: days -9, - 4, +3; NK: -4, -1 , +1 , +3, +6). Sorafenib (Sora, 90 mg/kg, p.o., daily) was injected for 8 days starting at day 0. Necropsy was performed at day 8 after treatment initiation (n = 4). (b) Flow cytometry analysis confirms the successful depletion of MF (F4/80⁺) and (c) NK cells (CD49b⁺). (d) Liver weights measured at day 8 after vehicle or sorafenib treatment indicate complete loss of sorafenib efficacy after depletion of MF and NK cells (e) Tumour lysates of all depletion groups were analyzed for cytokine secretion at day 8 of treatment with vehicle or sorafenib. Data indicate different cell type-specific cytokine releases. Differences between the groups were tested for significance using Student’s t-test (b - d) or one-way analysis of variance (ANOVA) followed by Tukey multiple comparison test (e) (^*p<0.05, ^**p<0.01 ,

^***p<0.001 ).

Figure 4: Depletion of macrophages reduces the therapeutic effect of sorafenib in Hep-55.1 c tumours, (a) Treatment schedule indicating the days of cell depletion (MF: days -9, -4 +3; NK/CD4/CD8: -4, -1 , +1 , +3, +6). Sorafenib (Sora, 90 mg/kg, p.o., daily) was injected p.o. for 8 days starting at day 0. Necropsy of mice of the control (Co) and sorafenib groups (n = 4 mice) was performed at day 8. (b) Weight of explanted HCC tumours of the iso antibody treated as well as MF, NK, NK/MF, CD4 and CD8 depletion groups at day 8 after vehicle or sorafenib treatment, respectively. Scatter plots (5 animals per group +/- SD) indicate a reduced sorafenib efficacy in MF and NK/MF depletion groups (c) Cytokine analysis of tumour lysates show a significant upregulation of pro-inflammatory cytokine secretion after 16 hours of treatment. Scatter plots present 4 tumours per group +/- SD. (b, c) Comparisons between groups were performed using the Student’s t-test (^*p<0.05,

^**p<0.01 , ^***p<0.001 ).

Figure 5: Sorafenib stimulates macrophages and NK cells and induces killing of Hep- 55.1c tumour cells, (a) Scheme illustrating the design of the co-culture experiment. Immune cells (MF, NK, NK/MF) were stimulated with DMSO or sorafenib (Sora, 10 mM) for 4 hours. The immune cells were transferred to the Hep-55.1c tumour cells, respectively, and were additionally treated with sorafenib or DMSO for 24 hours (b) Tumour cell killing determined by Hoechst33342/PI staining shows a significant induction of tumour cell death after co- culture of tumour cells with stimulated MF and NK cells. Bars are presented as fold induction of cell death normalized to cell death of DMSO-treated Hep-55.1 c cells (n = 3) +/- SD. (c) Supernatants of stimulated immune cells (MF, NK, NK/MF) as well as Hep-55.1c tumour cells were collected after 24 hours and were analyzed for cytokine secretion via multiplex protein assays. Data shown in scatter plots (n = 4) demonstrate the cell sources of different cytokines (d) Supernatants of Hep-55.1c tumour cells co-cultured with stimulated immune cells (MF, NK, NK/MF) for 24 hours were analyzed for cytokine secretion via multiplex protein assays. Data reveals an alternating cytokine release of different co-culture approaches (b - d) Differences between groups were tested for significance using one-way analysis of variance (ANOVA) followed by Tukey multiple comparison test (^*p<0.05,

^**p<0.01 , ^***p<0.001 ). (c - d) Values under detection level were marked by nd (not detected).

Figure 6: Sorafenib upregulates pro-inflammatory signatures and induces pyroptotic cell death of macrophages, (a) Heatmap showing selected genes from NanoString analysis of RNA samples isolated from iAST tumours treated p.o. with vehicle (Co) (n = 3) or sorafenib (Sora) (n = 2) for 16 hours. All genes depicted in the heatmap showed significant differences between groups determined by Student’s t-test. (b) Mice were treated with vehicle or sorafenib for 9 days and single tumour cells were analyzed via flow cytometry.

The percentages of MHC- tumour cells as well as the MHC-I expression levels (geometric mean) shown in scatter plots (n=4) +/- SD indicate a reduction of MHC-I on tumour cells (c) The caspase-1 activity after 4 hours treatment with DMSO, sorafenib, sunitinib, isotype control, anti-CSF-1 R and nigericin, respectively, analyzed in mouse (mu) and human (hum) MF, as well as Hep-55.1c tumour cells by in vitro bioluminescence assay indicates a sorafenib-triggered upregulation of caspase-1 restricted to MF. Bar charts show averages of treated cells (black) compared to averages of treated cells with additional caspase-1 inhibitor incubation (white) (n = 3) +/- SD. (b, c) Comparisons between groups were performed by Student’s t-test (^*p<0.05, ^***p<0.001 ).

Figure 7: Pyroptosis leads to macrophage cell death, (a) Analysis of caspase-1 activity after 24 hours of treatment with DMSO, sorafenib, sunitinib, isotype antibody, anti-CSF-1 R or nigericin, respectively, assessed by a bioluminescence assay in murine (mu) and human (hum) MF as well as Hep-55.1c tumour cells. A significant increase in caspase-1 activity was found in MF upon sorafenib and sunitinib stimulation. Nigericin was used as the positive control. Data reveal no significant change in caspase-1 activity in Hep-55.1 c tumour cells after treatment. Bar charts show averages of treated cells (black) compared to averages of treated cells with additional caspase-1 inhibitor incubation (dashed) (n = 3) +/- SD. (b) Cell death investigated by Hoechst33342/PI staining shows a significant increase in cell death in sorafenib-treated murine and human MF after 24 hours compared to DMSO-treated MF. Data are presented as averages (n = 3) +/- SD. (a, b) Differences between DMSO control and differently treated cells (black) were tested for significance using Student’s t-test (^*p<0.05, ^***p<0.001 ). Figure 8: Pyroptosis leads to macrophage cell death regardless of macrophage subtype, (a) Analysis of caspase-1 activity after 24 hours of treatment with DMSO, sorafenib, or nigericin, respectively, assessed by a bioluminescence assay in human (hum) M1 , M2 and M2c MF. A significant increase in caspase-1 activity was seen following sorafenib treatment for all three MF subtypes tested. Nigericin was used as a positive control. Differences between DMSO control and differently treated cells were tested for significance using Student’s t-test (^***p<0.001 ).

Figure 9: Hypothetic mode of action of pyroptotic macrophages inducing pro- inflammatory cytokine release and triggering NK cell cytotoxicity against HCC tumour cells. The therapeutic effect of sorafenib in HCC is induced by a stimulation of immune cells. The subsequent induction of pyroptosis in macrophages (MF) triggered by an upregulation of caspase-1 (step 1 ) leads to the release of pro-inflammatory cytokines (e.g. IL-18, IL1 B) (step 2). These cytokines induce the proliferation and activation of NK cells (step 3). The initiation of NK cell cytotoxicity and degranulation (step 4) provokes apoptosis of tumour cells (step 5). Besides this, sorafenib reduces the MHC-I expression of HCC tumour cells.

Figure 10: Macrophage and NK cell baseline in human HCC samples (a) shows the percentage of the area of 47 formalin-fixed, paraffin-embedded HCC tissue sections covered by macrophages. The average area covered is indicated (6.38%). (b) Shows the density of NK cells in the same 47 formalin-fixed, paraffin-embedded HCC tissue sections

(counts/mm²). The average density is again indicated (12.23 counts/mm²).

Detailed Description

The present inventors have shown that administration of a kinase inhibitor, such as sorafenib or sunitinib, to the patient induces pyroptosis of macrophages (TAMs) present in the tumour. TAM pyroptosis leads to the release of pro-inflammatory cytokines, thereby inducing NK cell proliferation, activation and tumour infiltration and tumour cell killing. In light of this, it is expected that kinase inhibitor treatment will be more effective in treating tumours comprising higher levels of macrophages, and optionally also higher levels of NK cells.

The present invention thus relates to methods of treating cancer in a patient, typically a human patient, wherein a tumour of the patient has, or has been determined to comprise, a level of macrophages above a predetermined threshold. The present invention also relates to a method of treating cancer in a patient comprising administering a therapeutically effective amount of a kinase inhibitor to a patient that has been determined to be responsive the kinase inhibitor based on the cancer comprising a macrophage level above a

predetermined threshold. A kinase inhibitor for use in a method of treating cancer in a patient, wherein a tumour of the patient has, or has been determined to comprise, a level of macrophages above a predetermined threshold similarly forms part of the present invention.

A method of treating cancer in a patient with a kinase inhibitor may comprise determining the level of macrophages in a tumour sample, such as a tumour biopsy, e.g. a tumour section, obtained from the patient, comparing the level of macrophages to a predetermined threshold, and treating a patient for whom the macrophage level exceeds the threshold with the kinase inhibitor. Alternatively, a method of treating cancer in a patient with a kinase inhibitor may comprise ordering test results determining the level of macrophages in a tumour sample obtained from the patient, and treating a patient for whom the macrophage level exceeds a predetermined threshold with the kinase inhibitor. Treatment preferably comprises administration of a therapeutically effective amount of the kinase inhibitor to the patient. The use of a kinase inhibitor for the manufacture of a medicament for the treatment of cancer in a patient, wherein a tumour of the patient has, or has been determined to comprise, a level of macrophages above a predetermined threshold is also contemplated.

In addition, a tumour of the patient may have, or have been determined to comprise, a level of NK cells above a predetermined threshold. Alternatively, a method of treating cancer in a patient, may further comprise determining the level of NK cells in the tumour sample obtained from the patient, comparing the level of NK cells to a predetermined threshold, and treating a patient for whom the NK cell level exceeds the threshold with the kinase inhibitor.

A method of determining the level of macrophages, and optionally NK cells, in a tumour sample obtained from a patient, and optionally comparing the macrophage level and NK cell level to a predetermined threshold is further provided. The method may be a method for providing information for predicting the response of a cancer patient to treatment with a kinase inhibitor.

The present invention also relates to a method of selecting a cancer patient for treatment with a kinase inhibitor, the method comprising determining the level of macrophages in a tumour sample, such as a tumour biopsy, e.g. a tumour section, obtained from the patient, comparing the level of macrophages to a predetermined threshold, and selecting the patient for treatment with the kinase inhibitor if the macrophage level exceeds the predetermined threshold. The method may further comprise determining the level of NK cells in the tumour sample obtained from the patient, comparing the level of NK cells to a predetermined threshold, and selecting the patient for treatment with the kinase inhibitor if the NK cell level exceeds the predetermined threshold.

The present invention also relates to methods of predicting the response of a cancer to treatment with a kinase inhibitor, the method comprising determining the level of

macrophages in a tumour sample obtained from the patient, and comparing the level of macrophages to a predetermined threshold, wherein a level of macrophages in the tumour sample which exceeds the threshold indicates that the cancer will, or is likely to, respond to treatment with the kinase inhibitor. The method may further comprise determining the level of NK cells in the tumour sample obtained from the patient, and comparing the level of NK cells to a predetermined threshold, wherein a level of NK cells in the tumour sample which exceeds the threshold indicates that the cancer will, or is likely to, respond to treatment with the kinase inhibitor. The advantage of these methods is that they can be performed prior to administration of the kinase inhibitor to the patient, thus ensuring that the kinase inhibitor is administered to patients likely to be benefit from the treatment, e.g. patient which are likely to show tumour growth inhibition or tumour regression in response to the treatment. A cancer which is likely to respond to treatment with a kinase inhibitor may have a probability of at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, or at least 90% of responding to treatment with the kinase inhibitor.

In instances where both the macrophage level and the NK cell level in a tumour sample is determined, the macrophage level and the NK cell level is each compared to a separate predetermined threshold. The determination of suitable thresholds is described herein.

The predetermined thresholds with which the macrophage level and optionally the NK cell level are compared are thresholds which are suitable for identifying tumours that are likely to respond to treatment with a kinase inhibitor, such as sorafenib, and may be determined in a number of ways as described below.

For example, the threshold with which the macrophage level is compared may be the mean, median, 60^th percentile, 65^th percentile, 70^th percentile, 75^th percentile, 80^th percentile, 85^th percentile, 90^th percentile, or 95^th percentile, preferably the 75^th percentile, of the distribution of macrophage levels in tumours of a cancer, such as HCC tumours. The mean, median,

60^th percentile, 65^th percentile, 70^th percentile, 75^th percentile, 80^th percentile, 85^th percentile, 90^th percentile, or 95^th percentile of the distribution of macrophage levels in tumours of a cancer may be determined based on the distribution of the macrophage levels in tumours of the cancer obtained from at least 500, at least 100, at least 50, or at least 25 individuals with said cancer.

Alternatively, the threshold with which the macrophage level is compared may be

determined using macrophage level data from at least 500, at least 100, at least 50, or at least 25 individuals with a cancer of interest, such as HCC, and network modelling to establish the threshold that allows maximal, optimal separation of cancer patients into responder and non-responder groups, whereby the responder group responds to treatment with the kinase inhibitor and the non-responder group does not respond to treatment with the kinase inhibitor.

The threshold with which the NK cell level is compared may be the mean, median, 60^th percentile, 65^th percentile, 70^th percentile, 75^th percentile, 80^th percentile, 85^th percentile, 90^th percentile, or 95^th percentile, preferably the 75^th percentile, of the distribution of NK cell levels in tumours of a cancer, such as HCC tumours. The mean, median, 60^th percentile, 65^th percentile, 70^th percentile, 75^th percentile, 80^th percentile, 85^th percentile, 90^th percentile, or 95^th percentile of the distribution of NK cell levels in tumours of a cancer may be determined based on the distribution of the NK cell levels in tumours of the cancer obtained from at least 500, at least 100, at least 50, or at least 25 individuals with said cancer.

Alternatively, the threshold with which the NK cell level is compared may be determined using NK cell level data from at least 500, at least 100, at least 50, or at least 25 individuals with a cancer of interest, such as HCC, and network modelling to establish the threshold that allows maximal, optimal separation of cancer patients into responder and non-responder groups, whereby the responder group responds to treatment with the kinase inhibitor and the non-responder group does not respond to treatment with the kinase inhibitor.

The patients from which the tumours used for macrophage and/or NK cell threshold determination are obtained are preferably patients which have not previously been treated with the kinase inhibitor in question, for example sorafenib. The patients from which the tumours used for macrophage and/or NK cell threshold determination are obtained may be patients which have not previously received treatment for the cancer.

The macrophage level detected is preferably the total macrophage level, e.g. the total TAM level. M1 macrophages are CD68+/CD163- and M2 macrophages are CD68+/CD163+. Methods of detecting macrophage levels in a tumour are known in the art and are described herein. For example, the macrophage level in a tumour may be detected by immunohistochemical staining of a tumour section using antibodies that bind CD163 and CD68, respectively. Suitable antibodies are known in the art. An exemplary

immunohistochemical method for detecting M1 and M2 macrophages is described in Barros et al. (2013). The macrophage level may the percentage of the area of the tumour section stained using such antibodies. For example, the 75^th percentile of the distribution of macrophage levels in HCC may be at least 5%, at least 6%, at least 7%, or at least 8% of HCC tumour section area covered by macrophages, i.e. HCC tumour sample section area stained using antibodies that bind CD163 and CD68.

Methods of detecting NK cell levels in a tumour are also known in the art and are described herein. NK cells are PRF1 +/CD3e-. In one example, the NK cell level in a tumour may be detected by immunohistochemical staining of a tumour section using antibodies that bind Prf1 and CD3e, respectively. Suitable antibodies are known in the art. The NK cell level may be the number of NK cells per area, e.g. per mm² of tumour section, as detected by immunohistochemical staining of a tumour section using such antibodies. For example, the 75^th percentile of the distribution of NK cell levels in HCC may at least 10, at least 1 1 , at least 12, at least 13, or at least 14 NK cells per mm² of HCC tumour sample section.

As an alternative to immunohistochemical staining, the macrophage level and/or NK cell level in a tumour may be detected by flow cytometry, for example. In this case, the macrophage and/or NK cell level may be the number of macrophages and/or NK cells per tumour volume, such as per cm³ of tumour sample.

The macrophage and/or NK cell level may be the ratio of macrophages and/or NK cells to the total number of immune cells present in the tumour. The total number of immune cells present may be detected by detection of CD45⁺ cells using e.g. an antibody that bind CD45.

Pyroptosis is known to be associated with the release of cytokines, including I L-1 b and IL-18 (Bergsbaken et al., 2009). In light of the finding by the present inventors that treatment with kinase inhibitors, such as sorafenib and sunitinib, induces pyroptosis of TAMs, as explained above, another possibility for predicting the response of a cancer patient to treatment with a kinase inhibitor involves the detection of IL-1 b and/or IL-18, e.g. in a sample, preferably a serum sample, obtained from the patient. In this case, detection is of IL-1 b and/or IL-18 may be performed before and after administration of the kinase inhibitor to the patient. Although such methods cannot be performed solely prior to the start of treatment with the kinase inhibitor, they have the advantage that the collection of serum samples for testing is less invasive than methods requiring tumour samples. The present invention thus relates to methods of predicting the response of a cancer to treatment with a kinase inhibitor, the method comprising administering the kinase inhibitor to the patient, determining the level of I L-1 b and/or IL-18, preferably IL-1 b and IL-18, in a sample, such as a serum sample, obtained from the patient, and comparing the level of IL- 1 b and/or IL-18 to a predetermined threshold, wherein a level of I L-1 b and/or IL-18 which exceeds the threshold indicates that the cancer will, or is likely to, respond to treatment with the kinase inhibitor. A method of determining the level of IL-1 b and/or IL-18 macrophages in a sample obtained from a patient, such as a serum sample, and optionally comparing the IL- 1 b and/or IL-18 level to a predetermined threshold is further provided. The method may be a method for providing information for predicting the response of a cancer patient to treatment with a kinase inhibitor.

A cancer which is likely to respond to treatment with a kinase inhibitor may have a probability of at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, or at least 90% of responding to treatment with the kinase inhibitor.

In instances where both the level of I L-1 b and the level of IL-18 is determined, the I L-1 b level and IL-18 level is each compared to a separate predetermined threshold. The determination of suitable thresholds is described herein.

The predetermined thresholds with which the IL-1 b level and the IL-18 levels are compared are thresholds which are suitable for identifying tumours that are likely to respond to treatment with a kinase inhibitor, such as sorafenib, and may be determined in a number of ways as described below.

For example the predetermined threshold may be the I L-1 b level and the IL-18 levels, respectively, in the patient prior to first administration of the kinase inhibitor to the patient. This is also referred to as the baseline level herein. Prior to treatment with the kinase inhibitor no macrophage pyroptosis should take place and hence the I L-1 b and IL-18 levels detected are expected to reflect the baseline levels of these cytokines in the patient in question. The predetermined threshold levels of I L-1 b and IL-18 may be determined in a sample, preferably a serum sample obtained from the patient.

In a preferred embodiment, the present invention thus relates to a method of predicting the response of a cancer to treatment with a kinase inhibitor, the method comprising (i) determining the baseline level (threshold) of I L-1 b and/or IL-18, preferably IL-1 b and IL-18, in a sample, such as a serum sample, obtained from the patient;

(ii) administering the kinase inhibitor to the patient;

(iii) determining the level of I L-1 b and/or IL-18, preferably I L-1 b and IL-18, in a sample, such as a serum sample obtained from the patient; and

comparing the level of I L-1 b and/or IL-18 determined in (iii) to the baseline level (threshold) of I L-1 b and/or IL-18 determined in (i), wherein a level of I L-1 b and/or IL-18 which exceeds the baseline level (threshold) indicates that the cancer will, or is likely to, respond to treatment with the kinase inhibitor.

Pyroptosis is a very fast form of cell death (~4h) followed by a release of the pyroptotic- specific cytokines (IL-1 b and IL-18) within 24-48 hours. It is therefore expected that elevated levels of I L-1 b and/or IL-18 will be detectable in patient serum soon after administration of the kinase inhibitor to the patient. Thus, the level of I L-1 b and/or IL-18 may be determined in the patient at least 24 hours after administration of the kinase inhibitor. The level of IL-1 b and/or IL-18 may be determined in the patient up to 36, or up to 48 hours after administration of the kinase inhibitor to the patient. For example, the level of I L-1 b and/or IL-18 may be determined about 24 hours after administration of the kinase inhibitor. Detection of I L-1 b and/or IL-18 soon after administration of the kinase inhibitor is advantageous as patients who are unlikely to benefit from the treatment can be identified quickly and ineffective treatment stopped prior to the onset of serious side effects associated with kinase inhibitor treatment.

A method of treatment as described may be comprise administering at least one further treatment to the patient in addition to the kinase inhibitor. The kinase inhibitor may thus be administered to a patient alone or in combination with one or more other treatments. Where the kinase inhibitor is administered to the patient in combination with another treatment, the additional treatment may be administered to the patient concurrently with, sequentially to, or separately from the administration of the kinase inhibitor. Where the additional treatment is administered concurrently with the kinase inhibitor, the kinase inhibitor and additional treatment may be administered to the individual as a combined preparation. The additional therapy may be a known therapy or therapeutic agent for the cancer to be treated.

The present inventors have shown that kinase inhibitors, such as sorafenib and sunitinib, mediate their effect on tumours at least in part by inducing pyroptosis of TAMs, leading to NK cell activation and tumour infiltration. In light of this, it is expected that administration of kinase inhibitors in combination with other immunotherapeutic agents may lead to enhanced efficacy. In particular, administration of kinase inhibitors with another agent or agents which induce NK cell activation is expected to be beneficial.

For example, the kinase inhibitor may be administered to the patient in combination with a checkpoint inhibitor, such as a PD-1 inhibitor. PD-1 is known to be expressed on NK cells. Inhibition of PD-1 may therefore further enhance NK cell activation. Examples of PD-1 inhibitors include anti-PD-1 antibodies, such as nivolumab and pembrolizumab.

Alternatively, the kinase inhibitor may be administered to the patient in combination with a checkpoint inhibitor, such as a CTLA-4 inhibitor. CTLA-4 mediates immunosuppression by e.g. inhibiting TReg cells and restoring T cell priming in the lymph node (Seidel et al., 2018). Examples of CTLA-4 inhibitors are known in the art and include ipilimumab and

tremelimumab.

Alternatively, the kinase inhibitor may be administered to the patient in combination with a toll-like receptor (TLR) agonist. TLRs are expressed on NK cells and TLR agonism may enhance NK cell activation and infiltration in the presence of kinase inhibitors.

As a further alternative, the kinase inhibitor may be administered to the patient in

combination with NK cells, in particular autologous NK cells. Without wishing to be bound by theory, it is thought that NK cell immunotherapy in combination with administration of a kinase inhibitor may lead to enhanced NK cell activation and tumour infiltration as a result of macrophage pyroptosis induced by the kinase inhibitor.

Kinase inhibitors will usually be administered in the form of a pharmaceutical composition, which may comprise at least one component in addition to the kinase inhibitor, such a pharmaceutically acceptable excipient, carrier, buffer, stabilizer or other materials well known to those skilled in the art.

Suitable routes for administration of kinase inhibitors to a patient are also known in the art. For example, a kinase inhibitor may be administered orally to a patient.

Administration of the kinase inhibitor may be in a "therapeutically effective amount", this being sufficient to show benefit to a patient. The actual amount administered, and rate and time-course of administration, will depend on the nature and severity of the cancer to be treated, the particular individual being treated, the clinical condition of the individual, the cause of the disorder, the method of administration, the scheduling of administration and other factors known to medical practitioners.

Suitable dosages for kinase inhibitors are known to medical practitioners. Prescription of treatment, e.g. decisions on dosage, frequency of administration etc., is within the responsibility of the medical practitioner and may depend on the severity of the symptoms and/or progression of the disease being treated, as well as the presence of side effects. For example, a typical dose for sorafenib is 800mg/day. A typical dose of sunitinib is 50mg/day. This is a dose for a single treatment of an adult individual, which may be proportionally adjusted for children and infants.

In patients with tumours that have been determined to comprise a level of macrophages, and optionally NK cells, above a predetermined threshold, a lower dose of the kinase inhibitor may be sufficient to provide a therapeutic effect than in patients whose tumours comprise lower levels of macrophages. Similarly, where the IL-1 b and/or IL-18 levels following administration of the kinase inhibitor to the patient are above a predetermined threshold a lower dose of the kinase inhibitor may subsequently be administered to the patient. Thus, the kinase inhibitor may be administered to the patient in a therapeutically effective amount, wherein the amount of the kinase inhibitor administered to the patient is at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, or at least 80% lower than the standard full dose of the kinase inhibitor administered to a patient, for example an adult human patient. The standard full dose of a given kinase inhibitor is known to a medical practitioner. In the case of sorafenib, the kinase inhibitor may for example be administered to the patient at a dose of no more than 600mg/day, no more than 400mg/day, or no more than 200mg/day. Treatment with kinase inhibitors such as sorafenib is associated with severe side-effects. Reducing the dose of the kinase inhibitor administered to the patient is expected to result in a reduction in the severity of such side effects. This has been shown in the context of sorafenib, where the dose is reduced to 600mg/day,

400mg/day, or 200mg/day if the patient experiences severe side effects to sorafenib treatment.

A patient as referred to herein is preferably a human patient. Where the method comprises determining the macrophage, and optionally the NK cell level, in a tumour of the patient, the patient has preferably not previously been treated with the kinase inhibitor, such sorafenib or sunitinib. The patient may be a patient which has had an inadequate response or intolerance to one or more immune checkpoint inhibitors. The determination of an inadequate response or intolerance to treatment with an immune checkpoint inhibitor is within the capabilities of the medical practitioner.

The kinase inhibitor as referred to herein is preferably a protein kinase inhibitor. The kinase inhibitor is a kinase inhibitor capable of inducing pyroptosis of macrophages, in particular TAMs, in a tumour of a patient. Thus, administration of the kinase inhibitor to the patient preferably induces pyroptosis of macrophages present in the tumour. Methods for determining the ability of a kinase inhibitor to induce macrophage pyroptosis are known in the art and are described herein. For example, pyroptosis of macrophages in the presence of a kinase inhibitor can be determined by detection of caspase-1.

Kinase inhibitors for use in cancer treatment are known in the art and include type II kinase inhibitors such as sorafenib, and type I kinase inhibitors such as sunitinib. Both sorafenib and sunitinib were shown by the present inventors to induce pyroptosis of TAMs, although sunitinib induced pyroptosis at a later time point after administration than sorafenib. The kinase inhibitor may therefore be sorafenib or sunitinib but preferably is sorafenib.

In a preferred embodiment, the kinase inhibitor is sorafenib, or a pharmaceutically acceptable salt, hydrate, or solvate thereof. More preferably, the kinase inhibitor is sorafenib tosylate, or a pharmaceutically acceptable hydrate or solvate thereof.

Sorafenib is co-marketed by Bayer and Onyx Pharmaceuticals as Nexavar (sorafenib tosylate). Alternatively, the kinase inhibitor may be sunitinib, or a pharmaceutically acceptable salt, hydrate, or solvate thereof. For example, the kinase inhibitor may be sunitinib malate, or a pharmaceutically hydrate or solvate thereof.

Sunitinib was previously known as SU11248 and is currently marketed by Pfizer as Sutent (sunitinib malate).

Alternative kinase inhibitors for use in the present invention include vandetanib, apatinib, or imatinib or a pharmaceutically acceptable salt, hydrate, or solvate thereof. For example, the kinase inhibitor may be apatinib mesylate or a pharmaceutically acceptable hydrate or solvate thereof.

The kinase inhibitor may be in the form of a salt of the kinase inhibitor, for example, a pharmaceutically-acceptable salt. Examples of pharmaceutically acceptable salts are discussed in Berge et ai, 1977,“Pharmaceutically Acceptable Salts,” J. Pharm. Sci.. Vol. 66, pp. 1-19.

For example, the substituted amino groups which may be cationic, and a salt may be formed with a suitable anion.

Examples of suitable inorganic anions include, but are not limited to, those derived from the following inorganic acids: hydrochloric, hydrobromic, hydroiodic, sulfuric, sulfurous, nitric, nitrous, phosphoric, and phosphorous.

Examples of suitable organic anions include, but are not limited to, those derived from the following organic acids: 2-acetyoxybenzoic, acetic, ascorbic, aspartic, benzoic,

camphorsulfonic, cinnamic, citric, edetic, ethanedisulfonic, ethanesulfonic, formic, fumaric, glucheptonic, gluconic, glutamic, glycolic, hydroxymaleic, hydroxynaphthalene carboxylic, isethionic, lactic, lactobionic, lauric, maleic, malic, methanesulfonic, mucic, oleic, oxalic, palmitic, pamoic, pantothenic, phenylacetic, phenylsulfonic, propionic, pyruvic, salicylic, stearic, succinic, sulfanilic, tartaric, toluenesulfonic, and valeric.

Examples of suitable polymeric organic anions include, but are not limited to, those derived from the following polymeric acids: tannic acid, carboxymethyl cellulose.

The kinase inhibitor may be in the form of a solvate of the kinase inhibitor, for example, a pharmaceutically-acceptable solvate. The term“solvate” is used herein in the conventional sense to refer to a complex of solute (e.g., compound, salt of compound) and solvent. If the solvent is water, the solvate may be conveniently referred to as a hydrate, for example, a hemi-hydrate, a mono-hydrate, a sesqui-hydrate, a di-hydrate, a tri-hydrate, etc. TAMs have been shown to be present in many solid human tumours. In view of the finding of the present inventors that kinase inhibitors, such as sorafenib and sunitinib, induce TAM pyroptosis and the release of pro-inflammatory cytokines, thereby inducing NK cell proliferation, activation and tumour infiltration and resulting in tumour cell killing, it is expected that tumours comprising macrophage, and optionally NK cell levels, above a predetermined threshold will respond to kinase inhibitor treatment. This include cancers that are known to be treatable with kinase inhibitors, such as HCC for example, as well as other solid cancers which are not currently treated with kinase inhibitors.

A cancer as referred to herein thus includes any type of solid cancer. The cancer is preferably responsive to treatment with a kinase inhibitor, e.g. type I or type II kinase inhibitors, such as sorafenib and/or sunitinib.

For example, the cancer may be a solid cancer selected from the group consisting of liver cancer (such as HCC), kidney cancer (such as RCC), thyroid cancer (such as DTC), gastrointestinal cancer (such as GIST), pancreatic cancer (such as neuroendocrine cancer of the pancreas), prostate cancer, lung cancer, bowel cancer, breast cancer, pancreatic cancer, oesophageal cancer, bladder cancer, brain cancer, stomach cancer, ovarian cancer, head and neck cancer, mesothelioma, uterine cancer, and melanoma. Preferably, the cancer is selected from the group consisting of HCC, RCC, DTC, gastrointestinal cancer, and neuroendocrine cancer of the pancreas, more preferably HCC, RCC and DTC. Most preferably the cancer is HCC. Cancers may be familial or sporadic. Cancers may be metastatic or non-metastatic.

Sorafenib has been shown to be suitable for the treatment of HCC, RCC and DTC. Thus, where the kinase inhibitor is sorafenib, the cancer may be HCC, RCC or DTC but preferably is HCC.

Sunitinib has been shown to be suitable for the treatment of RCC, gastrointestinal cancer, such as stromal tumours (GISTs), and neuroendocrine cancer of the pancreas. Thus, where the kinase inhibitor is sunitinib, the cancer may be RCC, gastrointestinal cancer, such as stromal tumours (GISTs), and neuroendocrine cancer of the pancreas.

Another aspect of the invention provides a kit for use in a method of predicting the response of a cancer to treatment with a kinase inhibitor. The kit preferably comprises components for detecting macrophages and optionally NK cells in a tumour sample obtained from a patient, such as antibodies. Antibodies suitable for the detection of macrophages include antibodies that bind CD163 and CD68, whilst antibodies suitable for the detection of NK cells include antibodies that bind perforin 1 (Prf1 ) and CD3e. M1 macrophages are CD68+/CD163- whilst M2 macrophages are CD68+/CD163+. NK cells are PRF1 +/CD3e-. In addition, or alternatively, the kit may comprise components for detecting I L-1 b and/or IL-18, such as antibodies which bind I L-1 b or IL-18, respectively. Suitable antibodies are known in the art and are described elsewhere herein. The components of the kit are preferably sterile and in sealed vials or other containers. A kit may further comprise instructions for use of the components in a method described herein. The components of the kit may be comprised or packaged in a container, for example a bag, box, jar, tin or blister pack.

It is to be understood that the application discloses all combinations of any of the above aspects and embodiments described above with each other, unless the context demands otherwise. Similarly, the application discloses all combinations of the preferred and/or optional features either singly or together with any of the other aspects, unless the context demands otherwise.

Other aspects and embodiments of the invention provide the aspects and embodiments described above with the term“comprising” replaced by the term“consisting of” and the aspects and embodiments described above with the term“comprising” replaced by the term ’’consisting essentially of”.

Modifications of the above embodiments, further embodiments and modifications thereof will be apparent to the skilled person on reading this disclosure, and as such, these are within the scope of the present invention.

“and/or” where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. For example“A and/or B” is to be taken as specific disclosure of each of (i) A, (ii) B and (iii) A and B, just as if each is set out individually herein.

Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the figures described above. Examples

1. Materials and Methods

7.7 Animal models.

For the iAST HCC tumour model, 6 to 8 week old female transgenic mice of Cre-inducible Albumin-specific SV40 T-antigen mouse model (iAST obtained from Charles River, Sulzfeld, Germany) with a body weight of 19-22 g were used (Stahl et al., 2009). HCC tumour growth was induced by intravenous (i.v.) injection of 5x10⁸ infectious units (IU) of adenovirus (Ad.Cre) expressing Cre recombinase (Vector BioLabs, Malvern, PA, USA). Due to the hepatocyte-specific albumin promoter of iAST mice, the multinodular tumour growth is restricted to the liver.

For the Hep-55.1c tumour model, C57BL/6 wildtype mice (Charles River) with an age of 8 to 9 weeks and a body weight of 20-23 g were used as donor animals. Mice were treated with the analgetics Metamizol (100 mg/kg; WdT, Garbsen, Germany) and Rimadyl (5 mg/kg; Zoetis, Berlin, Germany) prior to surgery as well as 24 and 48 hours after surgery. Animals were anesthetized with 2% isoflurane at 2 L/min oxygen and 5x10⁵ Hep-55.1c cells in 20 pl_ matrigel (Corning) were inoculated into the left lateral liver lobe. Abdomen and skin was closed with absorbable suture material (Prolene, Ethicon, Norderstedt, Germany). After 20 days orthotopic tumours of the donor animals were explanted and cut into fragments of 2 mm x 2 mm size under sterile conditions. One Hep-55.1c tumour fragment was implanted into the left lateral liver lobe of 8 week old female C57BL/6 mice by using a sterile trokar and fragment was fixed with 1 drop of degradable tissue glue (Histoacryl, B. Braun, Melsungen, Germany) to prepare the Hep-55.1 c mouse tumour model. Surgery of fragment implantation was performed analogously to intrahepatic cell inoculation as described above.

7.2 Micro-computed tomography (pCT).

Four hours prior to the first CT measurement mice were injected i.v. with 100 mI_ of the ExiTron nano 6000 contrast agent (Viscover, Miltenyi Biotec, Bergisch Gladbach, Germany). For the measurement mice were anesthetized with 2% isoflurane at 2 L/min oxygen and scanned with a dual-source pCT device (TomoScope Synergy Twin, CT Imaging GmbH, Erlangen, Germany) using a high resolution protocol (parameters: 1440 projections, tube voltage = 50 kV, tube current = 0.8 mA scan time = 180 s). Image reconstruction was performed using a cone-beam Feldkamp algorithm with a T50 reconstruction kernel and an isotropic voxel size of 35 pm. Images were visualized using the OsiriX software (Pixmeo, Bernex, Switzerland).

13 In vivo treatment

Mice were treated daily with 90 mg/kg of sorafenib tosylate (Nexavar; Bayer Schering Healthcare, Leverkusen, Germany) by oral gavage. Sorafenib was suspended in 7.5% Gelatine and 0.22% NaCI. Treatment was started at day 54 after virus injection for iAST mice and at day 35 after fragment implantation for Hep-55.1c mice, respectively. Murine IgG isotype control (clone MOPC-21 ; BioXCell) or anti-VEGF (B20-4.1 anti-murine/human VEGF-A) both at 10 mg/kg were administered intraperitoneally (i.p.) weekly. In vivo depletion

For macrophage (MF) depletion, mice were treated i.p. with 30 mg/kg anti-CSF-1 R (clone 2G2; Ries et al., 2014) on days -9 and -4 before first sorafenib therapy (“day 0”) as well as day +3 after sorafenib treatment initiation. For depletion of NK cells as well as CD4⁺ and CD8⁺ T cells, mice were treated i.p. with either 4 mg/kg anti-NK1.1 (clone PK136,

Biolegend), anti-CD4 (clone GK1.5, BioLegend) or anti-CD8a antibody (clone 53-6.7, BioXCell) on days -4 and -1 before sorafenib therapy and on days +1 , +3, +6 thereafter, respectively. Mouse lgG2a Isotype (clone MOPC-173, BioLegend) was injected i.p. as control at equal concentration. The depletion of cell populations was confirmed by flow cytometry using the following antibodies (clones): F4/80 (BM8) for (MF), CD49b (DX5) for NK cells, CD4 (RM4-5) for CD4⁺ T cells, CD8 (5H10) for CD8⁺ T cells and the matching isotype controls (BioLegend).

15 In vivo labelling of functional blood vessels.

In order to stain perfused blood vessels, mice received i.v. Alexa750-labeled lectin (100 pg, Bandeiraea simplicifolia BS-I, Sigma Aldrich) 5 minutes prior to necropsy. Mice were euthanized by cervical dislocation and tumours were explanted.

16 Flow cytometry

Tumours were explanted at the time points indicated in the figure legends. For flow cytometry analysis of multinodular iAST tumours, 3 nodules were pooled per mouse. In order to generate single-cell suspensions tumour samples were mechanically processed and digested with the enzymes DNAse I (0.01%; Roche) and collagenase IV (1 mg/ml; Sigma) at 37°C for 30 minutes by gentle shaking. Erythrocytes were lysed with lysing buffer (BD Biosciences) for 5 minutes at room temperature (RT). A cell number of 1x10⁶ cells per well was seeded in a v-bottom 96-well plate and Fc receptors were blocked with anti-mouse CD16/CD32 antibody (clone 2.4G2, BD Biosciences) for 5 minutes on ice. The following antibodies (clones) were used for immune cell staining: CD45 (30-F1 1 ), CD1 1 b (M1/70),

CD3 (17A2), CD4 (RM4-5), CD8a (53-6.7), F4/80 (BM8), NK1.1 (PK136), Ly6G (1A8), Ly6C (AL-21 ), MHC class II l-A/l-E (M5/1 14.15.2), CD1 1 c (N418), CD24 (M1/69), CD103 (2E7), CD206 (C068C2), CD69 (H1.2F3), CD279 (29F.1A12), CD107a (1 D4B), Ki67 (16A8) as well as appropriate isotype controls (all from BioLegend or BD Biosciences). Intracellular staining with Ki67 and CD107a antibodies was performed according to the manufacturer’s instructions of intracellular fixation and permeabilization kit (eBioscience). For CD107a degranulation analysis cells were stimulated with cell stimulation cocktail (eBioscience) for 1 hour followed by addition of Golgi Plug (BFA) & Stop (Monensin) medium (BD Biosciences) for 4 hours. DAPI (Roche) or fixable Zombie UV dye (BioLegend) was used to determine cell viability. Measurements were performed using the LSRFortessa device (BD Biosciences) and data were analyzed using FlowJo software version 10 (Treestar). Cells were initially gated to define single cells and discriminate between live and dead cells. The different immune cell subsets were identified as follows: total immune infiltrate as CD45⁺; CD4⁺ T cells as CD45⁺CD1 1 b CD3⁺CD4⁺; CD8⁺ T cells as CD45⁺CD1 1 b CD3⁺CD8a⁺; dendritic cells (DC) as CD45⁺Ly6C Ly6G MHC class II⁺F4/80^|OWCD24-; monocytic MDSC as CD45⁺CD1 1 b⁺ F4/80 Ly6C^hi9h Ly6G^_; granulocytic MDSC as CD45⁺CD1 1 b⁺F4/80 Ly6C^lowLy6G⁺; MF as CD45⁺CD1 1 b⁺F4/80⁺; M1 MF as CD45⁺CD1 1 b⁺F4/80⁺ CD1 1 c^h'^9hCD206^low; M2 MF as CD45⁺CD1 1 b⁺F4/80⁺ CD1 1 c^lowCD206^hi9h; NK cells as CD45⁺NK1 .1 ⁺.

17 Histology

For histological analysis, formalin fixed paraffin embedded tissue samples (FFPET) were cut into 1.5 pm thin sections. Slides were deparaffinized in a descending xylene and ethanol series and rehydrated in deionized water for 30 s. Afterwards, antigen retrieval and protein blocking (Dako) were conducted. Various primary antibodies were used for

immunohistochemical (IHC) and immunofluorescence (IF) stainings: Ki67 (30-9, Ventana Medical systems), F4/80 (BM8, Acris), cleaved caspase-3 (ASP175, Cell signaling), CD31 (clone SZ31 , Dianova), osmooth muscle actin (a-SMA) (clone 1A4, Sigma Aldrich), CD3 (clone SP7, Spring Bio). IHC staining for Ki67 and F4/80 were performed by using the Benchmark XT automated stainer (Ventana Medical systems). The primary antibody was captured by a secondary OmniMap antirabbit HRP antibody and signals were coloured with DAB and counterstained via hematoxylin. IF staining for cleaved caspase-3, CD31 , a-SMA and CD3 were performed by an incubation step with the primary antibody (1 :100) for 1 hour followed by incubation with fluorescent-labeled secondary antibody (Thermo Fisher) for 30 min in dark. Subsequently, the sections were covered with a DAPI-containing mounting medium (Fluoro-Gel II, Electron Microscopy Sciences, Hatfield, USA) and locked with a cover slide. All brightfield and fluorescence tissue slides were scanned with the AxioScan device (Zeiss, Germany). Image visualization was performed using the corresponding slide scanner software ZEN (Zeiss, Germany). Positive signals were quantified by HALO software (Indica Labs, Corrales, NM, USA).

1.8 Cytokine detection

In order to analyze cytokine levels, tumours were explanted at time points indicated in the figure legends. Samples were frozen in liquid nitrogen and homogenized by using Precellys ceramic bead tubes (1.4 mm bead size) as well as tissue lysis buffer (Biorad) containing PMSF, factor 1 and 2. Protein concentration was determined by Pierce bicinchoninic acid protein assay kit (Thermo Fisher). Cytokines were detected by multiplex BioPlex Pro kit (mouse cytokine 23-plex and 9-plex, Biorad). Amounts of 40 pg tumour lysates and 50 pL cell culture supernatants were used, respectively, and analyzed on the BioPlex 200 instrument (Biorad).

1.9 Cell culture

Hep-55.1c tumour cells were cultured in T75 flasks with DMEM high glucose medium (PAN Biotech) supplemented with fetal calf serum (10%, Gibco) and L-glutamine (5%, PAN Biotech).

1.10 NK cell isolation

NK cells were isolated from mouse spleens using a NK cell isolation kit (Miltenyi Biotech, Bergisch Gladbach, Germany) according to manufacturer’s instructions. Isolated NK cells were cultured with RPMI-1640 medium (PAN Biotech) supplemented with fetal calf serum (10%; Gibco), L-glutamine (5%; PAN Biotech), NEAA (5%; Thermo Fisher) and PenStrep (1 %; Gibco).

1.11 Monocyte differentiation

Mouse femurs were explanted under sterile conditions and bone marrow (BM) containing monocytes was flushed out with PBS. Cells were centrifuged and red blood cells were lysed with lysis buffer (BD Biosciences) for 5 minutes at RT. BM cells were cultured in PermaLife bags (Origene) using RPMI-1640 medium (PAN Biotech) supplemented with fetal calf serum (10%; Gibco), L-glutamine (5%; PAN Biotech), non-essential amino acids (NEAA; 5%, Thermo Fisher), PenStrep (1%; Gibco) and murine colony-stimulating factor 1 (CSF1 , 20 ng/mL, R&D) for 6 days to generate M2 macrophages. Human monocytes were isolated from healthy donors and cultured in PermaLife bags using RPMI-1640 medium

supplemented with fetal calf serum (10%), L-glutamine (5%), NEAA (5%), PenStrep (1%). Human CSF1 (20 ng/mL, Biomol) and human CSF2 (50 ng/ml, Biomol) was added to the medium for 6 days to generate M2 or M1 macrophages, respectively. In order to differentiate M2c macrophages, human IL10 (10 ng/ml, R&D) was added for additional 24 hours to M2 macrophages.

172 Co-culture experiment

Hep-55.1 c tumour cells were labelled with the CellTracker dye (1 mM CMFDA, Life

Technologies) and seeded in flat-bottom 96-well plates (5x10³ cells per well). Hep-55.1 c cells were used in this experiment as iAST tumour cells have been found too fragile for the required cell sorting procedure. NK cells and MF, respectively, as well as a mix of NK cells and MF, were seeded in u-bottom 96-well plates (5x10³ cells per well) and treated with sorafenib (10 pM in DMSO) for 4 hours followed by medium exchange to remove excess of sorafenib. The following day, the pre-treated NK cells, MF and co-cultured NK cells and MF were transferred onto the tumour cells. Subsequently, co-cultured tumour and immune cells were treated with sorafenib tosylate (10 pM) or vehicle control (DMSO) for another 24 hours. Prior to imaging, the cells were stained with 5 pg/mL Hoechst3342 (Invitrogen) and 5 pg/mL propidium iodide (PI) (Invitrogen) to discriminate between live and dead cells. Images were acquired with an Operetta imaging system (Perkin Elmer) using a 10x objective. Each sample condition was performed in triplicates and 9 fields per well were scanned.

1.13 Caspase-glo 1 assay

Cells (mouse M2 MF, human M1/M2/M2c MF and Hep-55.1c) were seeded in 96-well plates (2x10⁴ cells in 100 pL medium/well) and treated with sorafenib tosylate (10 pM), sunitinib malate (10 pM, Sigma), murine lgG1 isotype (10 pg/ml, BioXCell), anti-mouse-CSF- 1 R (10 pg/ml, clone 2G2), human lgG1 isotype (10 pg/ml, in-house), anti-human CSF-1 R (10 pg/ml, emactuzumab; Ries et al., 2014), nigericin (10 pM, Invivogen) for 4 and 24 hours. Assay reagents of the caspase-glo 1 kit (Promega, Fitchburg, Wl, USA) were prepared according to manufacturer’s instructions. 100 pL of the assay reagent was added to half of the wells, while the other half were loaded with 100 pL of the Ac-YVAD-CHO inhibitor reagent (1 pM) in order to measure caspase-1 -independent peptide cleavage. After 1 hour incubation at RT, luminescence was measured by a microplate reader (Tecan Infinite 2000).

1.14 NanoString analysis.

Tumour tissues were taken 16 hours after sorafenib therapy and directly frozen in liquid nitrogen. Frozen tissues were homogenized in RNA lysis buffer followed by RNA isolation with RNEasy Mini kit (Qiagen). Two panels of the NanoString nCounter platform (NanoString Technologies) were used (Mouse Immunology Panel for 561 genes, Mouse Myeloid Innate Immunity Panel for 754 genes) to analyze gene expression in tumour RNA samples (200 ng per sample). The code set was hybridized with the RNA at 65°C overnight. RNA transcripts were immobilized and counted using the NanoString nCounter Digital Analyzer. Normalized raw expression data were analyzed when 2 SDs above the geometric mean of the codeset- internal negative control probes were reached. Genes were excluded from further analysis if 90% of their expression was below the background threshold. Genes were normalized to the geometric mean of the internal positive controls as well as to 4 housekeeping genes (Eefl g, RPL19, Ppia and Tbp for the Mouse Immunology Panel and Eefl g, Polrl b, Ppia and Sf3a3 for the Mouse Myeloid Innate Immunity Panel).

1.15 Sta tistical analysis

For statistical analysis, GraphPad Prism software version 6 (GraphPad Software Inc., San Diego, CA, USA) was used. Graphs represent mean values and error bars indicate standard deviations (SD). The number of replicates is indicated in the figure legends. Differences between 2 groups were tested for significance using the Student’s t-test. Statistical comparisons between 3 or more groups were performed using one-way analysis of variance (ANOVA) followed by Tukey multiple comparison test. P values less than 0.05 were considered statistically significant (^* = p < 0.05, ^** = p < 0.01 , ^*** = p < 0.001 ).

1.16 IHC tissue analysis and macrophage and NK cell level cut-off definition

2.5 pm thick formalin-fixed, paraffin-embedded tissue sections of 47 human HCC samples were used for CD163/CD68 staining to detect macrophages and Prf1/CD3e staining to detect NK cells. The stainings were performed using a BenchMark XT automated stainer with the NEXES version 10.6 software and the following reagents were used: CD163 Mouse MRQ-26 IgG (Incubation Time = 8 min, Lot 1620803B, Ventana Medical Systems), CD68 Mouse PG-M1 lgG3 (Incubation Time = 60 min, Lot 20029531 , DAKO), Perforin 1 Mouse 5B10 lgG1 (Incubation Time = 12min, Lot G3203772, Abeam), CD3e Rabbit 2GV6

(Incubation Time = 16 min, Lot Y26143, Ventana Medical Systems). Stained slides were scanned at 20x using the iScan high throughput scanner (Ventana) for automated digital analyses (Chen and Srinivas, 2015) to detect and quantify macrophage populations

(CD68+/CD163+; CD68+/CD163-) as well as NK cells (PRF1 +/CD3e-) in tissues. Scanned images were processed to obtain single stain channel images to separate the specific staining of the biomarker from that of hematoxylin, followed by the detection and

identification of objects of interest. The detected objects were then used to compute the area of coverage in the case of macrophages as well as the counts/mm² in the case of NK cells to determine the baseline levels of macrophages and NK cells in HCC prior to sorafenib treatment (Fig. 10a, b). Example 1 - Sorafenib leads to reduced HCC tumour load

To investigate the effects of sorafenib on HCC tumours, first a single injection of adenovirus expressing the Cre recombinase (Ad.Cre) was performed to iAST mice inducing a multinodular tumour growth in the liver after 7 to 8 weeks. From day 54 after virus injection, sorafenib or vehicle were administered daily (Fig. 1a) and tumour sizes were measured by micro-computed tomography (pCT) over 8 days. Mice treated with sorafenib tosylate showed a strong tumour growth inhibition compared to vehicle-treated mice, which is in line with previous reports (Runge et al., 2014). The effects seen by pCT were supported by visual examination of the explanted livers and reflected by lower liver and body weights of treated animals (Fig. 1 b). Interestingly, the strong decrease of tumour load by sorafenib could not be accounted to changes in tumour cell proliferation determined by Ki67 staining of HCC sections (Fig. 1 c). However, caspase-3 expression as an indicator for apoptosis was clearly upregulated in the treated HCC tumours (Fig. 1d).

As second HCC model, Hep-55.1c tumour fragments were implanted into the left lateral liver lobe of C57BL/6 mice. Also in this orthotopic model, sorafenib treatment decreased tumour progression (Fig. 4b), and explanted livers displayed a smaller tumour load compared to controls (Fig. 4c), but to a lower extent compared to iAST tumours.

Example 2 - Efficacy of sorafenib is not mediated bv antianaioaenic effect.

Next, the antiangiogenic capacity of sorafenib was assessed in both orthotopic HCC models. For this purpose, mice were treated for 10 days with either sorafenib or anti-VEGF antibody. The vessel architecture was studied histologically after in vivo lectin application to mark functional vessels as well as ex vivo staining for CD31 (all vessels) and a-SMA (mature vessels). The total vessel density (CD31 ⁺) as well as the fraction of functional vessels (Lectin⁺) were significantly reduced by both, sorafenib and anti-VEGF. In addition, each monotherapy resulted in a strong reduction of pericyte coverage (a-SMA⁺) of endothelial cells. Interestingly, in contrast to the strong inhibition of tumour progression by sorafenib in iAST mice, the anti-VEGF treatment had no effect on tumour growth as indicated by pCT and the weight of explanted livers. However, despite these similar vascular responses only sorafenib inhibited ultimately tumour progression in iAST mice. In the Hep-55.1c model, the vessel density at baseline was low. Here, sorafenib treatment over 10 days did not alter the fraction of endothelial cells (CD31⁺) or the pericyte coverage (a-SMA⁺) despite tumours regression. These results indicate that alternative therapeutic mechanisms of sorafenib mediate the anti-tumoural activity of sorafenib in both HCC models. Example 3 - Sorafenib modulates the intra-tumoural immune infiltrate towards a pro- inflammatory milieu

To analyze the immune-related mechanisms contributing to the efficacy of sorafenib, the immune cell composition of iAST tumours after treatment with vehicle or sorafenib tosylate was characterised by flow cytometry. MF density was decreased by sorafenib therapy (Fig. 2a), which was further confirmed by immunohistochemical analysis of F4/80 stained tumour sections (Fig. 2b). The predominant M1 -like polarization based on the expression of the markers CD1 1 c^hi9hCD206^low within the F4/80⁺ population remained unchanged (Fig. 2a). In addition, the total number of NK cells, their activation state (CD69 and PD-1 ) and

degranulation activity (CD107a) was strongly increased by sorafenib. In accordance with these findings, elevated levels of Ki67⁺ NK cells were found after sorafenib treatment (Fig. 2c). Next, the T cells within the iAST tumours were analysed. Flow cytometry and IHC analysis revealed that CD4⁺ and CD8⁺ T cell populations remained unchanged upon sorafenib treatment (Fig. 2d,e). Furthermore, no significant changes were found for the overall CD45⁺ total immune infiltrate, dendritic cells (DC) as well as monocytic and granulocytic myeloid derived suppressor cells (MDSC, Fig. 2f). To gain further insights into sorafenib-triggered immunological changes, the cytokine and chemokine profiles after 16 hours treatment were analysed. Compared to vehicle-treated tumours, sorafenib increased the protein expression levels of several pro-inflammatory cytokines involved in innate immune responses such as IL1A (IL-1 a), IL1 B (IL-1 b), IL6, IL12, IL15, IL-18, IFNG (IFNy) and CSF2 (Fig. 2g). Only TNF expression was not significantly increased. Levels of CCL5, which supports the recruitment of lymphocytes to the inflammatory sites, were strongly upregulated upon sorafenib treatment. Analogous experiments in the Hep55.1 c model gained comparable results (Fig. 4d). Here, the chemokine CCL3, that attracts macrophages, was significantly higher expressed. Taken together, these data suggest that the therapy of HCC tumours with sorafenib induces a profound pro-inflammatory immune response in the tumour microenvironment.

Example 4 - Tumour growth inhibition bv sorafenib is dependent on macrophages and NK cells

Encouraged by the flow cytometry data, the functional role of the NK cells and MF in sorafenib tosylate-treated tumours was investigated. In order to dissect the influence of MF and NK cells, MF were depleted with an anti-CSF1 R antibody and NK cells with an anti- NK1.1 antibody individually and in combination (Fig. 3a). Successful depletion of immune cell subsets was confirmed by flow cytometry at day 1 of sorafenib treatment (Fig. 3b, c). Pre- treatment with the IgG control antibody did not interfere with sorafenib’s therapeutic efficacy (Fig. 3d). In contrast, the depletion of MF completely abolished the tumour growth inhibition effect of sorafenib (Fig. 3b). The depletion of NK cells (NK^_), resulted in a reduced therapeutic efficacy of sorafenib as well. The depletion of both, MF and NK cells (NK7MF-), clearly increased tumour progression in the iAST model (average weight of explanted livers plus tumour: 5.2 +/- 1.8 g (IgG), 7.1 +/- 0.9 g (NK /MF )) (Fig. 3d). Images of explanted livers (Fig. 3e) confirmed the finding that MF are fundamental mediators of HCC tumour reduction triggered by sorafenib.

In addition, it was confirmed that T cells do not play a role in the therapeutic effect of sorafenib in iAST mice by also depleting these cells prior to sorafenib treatment with either anti-CD4 or CD8 antibodies. Neither CD4⁺ nor CD8⁺ T cell depletion had an effect on sorafenib-induced tumour reduction shown by liver weight and liver images taken at treatment day 8.

The depletion experiments were also performed in the Hep-55.1 c tumour model. Here, a moderate effect on tumour load was detected (Fig. 4b). As in the iAST model, MF depletion abrogated the anti-tumour efficacy of sorafenib while NK cell depletion slightly reduced the sorafenib-triggered tumour reduction. The NK /MF cohort showed similar sorafenib-induced tumour growth inhibition as the single MF depletion cohort (MF ). No significant changes in sorafenib response were found in Hep-55.1 c tumours after CD4⁺ and CD8⁺ T cell depletion (Fig. 4), demonstrating that the immune-mediated response to sorafenib is also not T cell dependent in this model.

Next, the cytokine expression profile in tumour lysates of the different depletion groups (MF-, NK- or NK /MF ) was analysed and compared with the cytokine expression profile in the control group (Fig. 3f). Multiple cytokines (I L1 A, IL1 B, IL12, IL15, IL-18, TNF) were upregulated upon sorafenib treatment. MF depletion again abolished these effects, indicating that these cytokines were released by MF. In addition, some cytokines that were found to be elevated by sorafenib, such as IFNG and CSF2, were reduced after NK cell depletion, indicating that these cytokines were released by NK cells.

Taken together, the results of the depletion study, together with the cytokine expression analysis, strongly suggest a close interplay between MF and NK cells that is mediated by pro-inflammatory cytokines following sorafenib treatment, thereby promoting an innate anti- tumour immune response. Example 5 - Sorafenib stimulates the interplay of macrophages and NK cells required for

HCC tumour cell killing

To delineate in more detail the interaction between MF, NK and tumour cells, in vitro co- culture studies were performed. For this purpose, immune cells (MF, NK cells and co-culture of MF and NK cells) were isolated and pre-treated with either sorafenib or DMSO for 4 hours (Fig. 5a). In parallel, CMFDA-labeled Hep-55.1 c tumour cells were seeded in 96-well plates. After 24 hours, the immune cells were added to the Hep-55.1 c tumour cells and additionally treated with either DMSO or sorafenib, respectively, for further 24 hours. Prior to imaging, cells were stained with Hoechst33342/PI and the fraction of dead tumour cells (PI7CMFDA⁺) was determined by imaging. Induction of Hep-55.1 c cell death after treatment with sorafenib did not significantly differ from DMSO-treated tumour cells. Co-culturing with either MF or NK cells followed by sorafenib therapy did not strongly increase tumour cell death. In contrast, sorafenib pre-treatment of both, MF and NK cells, caused a 4-fold increase of Hep- 55.1 c tumour cell death after co-culturing with both immune cell populations (Fig. 5b).

Next, cytokine secretion by the immune cells, as well as Hep-55.1 c tumour cells, into the culture supernatant after sorafenib treatment was analysed. Particularly MF showed an enhanced release of pro-inflammatory cytokines and chemokines (I L1 A, IL6, IL12B, IL12, IL- 18, CCL3, CCL5; Fig. 5c) that was maintained in the co-culture with Hep-55.1 c tumour cells (Fig. 5d). The cytokine CSF2 was the only exception, as it was exclusively expressed in the tumour co-culture setting. Importantly, both in vitro and in vivo cytokine-/chemokine profiles of sorafenib treatment matched very well. These in vitro experiments support the in vivo observations, and strengthen the hypothesis that the interplay between MF and NK cells is triggered by sorafenib and then promotes the killing of Hep-55.1 c tumour cells. Cytokine release of DMSO-treated immune cells and tumour cells was not detectable.

Example 6 - Sorafenib regulates a pro-inflammatory gene signature, downreaulates MHC class I on tumour cells and induces pyroptotic death of macrophages

In order to delineate the immune signaling pathways regulated by sorafenib, NanoString analyses of iAST tumour RNA samples after 16 hours treatment with either vehicle or sorafenib tosylate were performed. 23 genes displayed significantly up- or downregulated expression following sorafenib treatment and these are displayed in a heatmap (Fig. 6a). Among the sorafenib-upregulated genes pro-inflammatory-related (IL1 B, IL6, IL23, IL33) and innate immune response-related genes (CR2, FCGR1 , LILRA5, STING) were identified. Notably, also genes involved in immune cell chemotaxis were activated following sorafenib treatment (CCL3, CCL24, CCR5, CXCL14). Moreover, an increase in TLR6, PDGFRA, RAB20, CSF1 , CSF1 R levels was seen. With regard to NFKB signaling, an upregulation of CMKLR1 and TRAF4 was detected, confirming previously reported results by Sprinzl et al. (2013). mRNA analysis also demonstrated decreased levels of CD34, a gene associated with T cell adhesion and migration. Interestingly, sorafenib-treated tumours were

characterized by downregulation of histocompatibility class II antigens (H2-Aa, H2-Eb1 ), suggesting a decrease in antigen presentation. Corresponding to these results a substantial downregulation of MHC class I was detected by flow cytometry on the iAST tumour cells, as well as reduced frequencies of MHC class G cells (Fig 6b).

Finally, the Nanostring analysis also uncovered enhanced mRNA levels of caspase-1 (CASP1 ) most likely accounting for cell death induced by sorafenib tosylate treatment (Fig. 6a). Based on these results, a caspase-1 detection assay was performed to determine whether pyroptosis is involved using murine and human MF as well as Hep-55.1c tumour cells (Fig. 6c). A significant increase in caspase-1 activity at 4 hours sorafenib treatment was detected in both, murine and human MF (Fig. 6c) with nigericin serving as positive control for pyroptosis. Pyroptosis by sorafenib and nigericin in MF was specifically inhibited by addition of the Ac-YVAD-CHO caspase-1 inhibitor. In order to check whether pyroptosis of MF is a sorafenib-specific mechanism, additional therapeutic agents (sunitinib malate, anti- CSF-1 R and the corresponding isotype antibody) were used but were found not to result in elevation of caspase-1 activation (Fig. 6c). Caspase-1 activity in Hep-55.1 c tumour cells remained unchanged confirming that sorafenib specifically upregulates caspase-1 in MF and induces pyroptotic cell death. While sorafenib significantly increased caspase-1 activity already after 4 hours, sunitinib only showed a slight increase in murine and human MF after 24 hours incubation (Fig. 7a). Notably, pyroptosis of MF or Hep-55.1 c tumour cells stimulated with anti-CSF-1 R or the corresponding isotype antibody was not detected at any time point (Fig. 6c, Fig. 7a). The cell death analysis via Hoechst33342/PI staining revealed that murine and human MF die after 24 hours treatment with sorafenib, whereas Hep-55.1 c tumour cells are not affected (Fig. 7b).

Example 7 - Sorafenib induces pyroptosis of macrophages independent of macrophage subtype

In order to determine whether the pyroptotic effect of sorafenib tosylate is dependent on the macrophage subtype, a caspase-1 detection assay was performed (Fig. 8). A significant increase in caspase-1 activity after 4 hours sorafenib treatment was detected in all three macrophage subsets (M1 , M2, M2c). Nigericin served as positive control for pyroptosis. Example 8 - Determination of macrophage and NK cell base-line levels in HCC Analysis of 47 human HCC samples from patients who have not previously been subjected to a therapy for HCC for the presence of macrophages (CD68+/CD163+ [M2] and

CD68+/CD163- [M1 ]) and NK cells (Prf1 +/CD3e-) showed that HCCs vary widely in the level of macrophages and NK cells they contain, with only few HCCs comprising high levels of macrophages and/or NK cells (Fig. 10). This is consistent with the observation that only a small percentage of HCC patients respond to sorafenib treatment.

The threshold for a high macrophage level was defined as the 75^th percentile of the distribution of macrophage levels detected in 47 human HCC tissue sections. The macrophage level detected was the percentage of each HCC tissue section area covered by macrophages, with the 75^th percentile being equivalent to 8.3% of tissue section area covered by CD68+/CD163+ (M2) and CD68+/CD163- (M1 ) macrophages (Figure 10a). The threshold for a high NK cell level was similarly defined as the 75^th percentile of the distribution of NK cell levels detected in 47 human HCC tissue sections. The NK cell level detected was the number of NK cells/mm² of tumour section, with the 75^th percentile being equivalent to 14.7 counts of PRF1 +/CD3e- NK cells per mm² of tissue section (Figure 10b). Based on the thresholds defined above, 12 out of 47 patients (25.5%) had high intratumoral macrophage levels and 12 out of 47 patients (25.5%) had high intratumoral NK levels, while 3 out of 47 patients (6.38%) had both high intratumoral macrophage and NK cell levels, showing that high intratumoural macrophage and NK cell levels is particularly a rare in HCC patients. This consistent with the very small percentage of HCC patients that show a complete response to sorafenib treatment.

References

All documents and sequence database entries mentioned in this specification are

incorporated herein by reference in their entirety for all purposes.

Bergsbaken et al. Pyroptosis: host cell death and inflammation. Nat Rev Microbiol. 7(2): 99-109 (2009). Branco, F. et al. The Impact of Early Dermatologic Events in the

Survival of Patients with Hepatocellular Carcinoma Treated with Sorafenib. Ann Hepatol. 16, 263-268 (2017).

Barros, M.H., et al., Macrophage polarisation: an immunohistochemical approach for identifying M1 and M2 macrophages. PLoS One, 2013. 8(11 ): p. e80908.

Cho et al. Clinical parameters predictive of outcomes in sorafenib-treated patients with advanced hepatocellular carcinoma. Liver International 950-957 (2013)

Ferlay, J. et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136, 359-86 (2015).

Hayashi et al. Serum cytokine profiles predict survival benefits in patients with advanced hepatocellular carcinoma treated with sorafenib: a retrospective cohort study.

BMC Cancer (2017) 17:870.

Janevska, D., Chaloska-lvanova, V., Janevski, V. Hepatocellular Carcinoma: Risk Factors, Diagnosis and Treatment. Open Access Maced J Med Sci. 3, 732-6 (2015).

Keating, G. M. Sorafenib: A Review in Hepatocellular Carcinoma. Target Oncol. 12, 243-253 (2017).

Llovet, J. M. et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 359, 378-90 (2008).

Reig, M. et al. Early dermatologic adverse events predict better outcome in HCC patients treated with sorafenib. J Hepatol. 61 , 318-24 (2014).

Ries, C. H. et al. Targeting tumour-associated macrophages with anti-CSF-1 R antibody reveals a strategy for cancer therapy. Cancer Cell 25, 846-59 (2014).

Rimola, J. et al. Complete response under sorafenib in patients with hepatocellular carcinoma: Relationship with dermatologic adverse events. Hepatology 10, 1002 (2017).

Runge, A. et al. An inducible hepatocellular carcinoma model for preclinical evaluation of antiangiogenic therapy in adult mice. Cancer Res. 74, 4157-69 (2014).

Seidel et al., Anti-PD-1 and Anti-CTLA-4 Therapies in Cancer: Mechanisms of Action, efficacy, and Limitations. Volume 8, Article 86 (2018).

Sprinzl, M. F. et al. Sorafenib perpetuates cellular anticancer effector functions by modulating the crosstalk between macrophages and natural killer cells. Hepatology 57, 2358-68 (2013). Stahl, S. et al. Tumour agonist peptides break tolerance and elicit effective CTL responses in an inducible mouse model of hepatocellular carcinoma. Immunol Lett. 123, 31- 7 (2009).

Wilhelm, S. M., Adnane L., Newell P., Villanueva A., Llovet J. M., Lynch M.

Preclinical overview of sorafenib, a multikinase inhibitor that targets both Raf and VEGF and PDGF receptor tyrosine kinase signaling. Mol Cancer Ther. 7, 3129-40 (2008).

Yada, M. et al. Indicators of sorafenib efficacy in patients with advanced

hepatocellular carcinoma. World J Gastroenterol. 20, 12581-7 (2014).

Claims

Claims

1. A kinase inhibitor for use in a method of treating cancer in a patient, wherein a tumour of the patient has been determined to comprise a level of macrophages above a predetermined threshold.

2. A kinase inhibitor for use according to claim 1 , wherein the tumour has further been determined to comprise a level of natural killer (NK) cells above a predetermined threshold.

3. A kinase inhibitor for use in a method of treating cancer in a patient, the method comprising

determining the level of macrophages in a tumour sample obtained from the patient; comparing the level of macrophages to a predetermined threshold; and

treating a patient for whom the macrophage level exceeds the threshold with the kinase inhibitor.

4. A method of treating cancer in a patient with a kinase inhibitor, wherein the method comprises:

determining the level of macrophages in a tumour sample obtained from the patient; comparing the level of macrophages to a predetermined threshold; and

administering a therapeutically effective amount of the kinase inhibitor to a patient for whom the macrophage level exceeds the threshold.

5. A kinase inhibitor for use according to claim 3 or a method according to claim 4, wherein the method further comprises

determining the level of NK cells in the tumour sample obtained from the patient; comparing the level of NK cells to a predetermined threshold; and

treating a patient for whom the NK cell level exceeds the threshold with the kinase inhibitor.

6. A method of predicting the response of a cancer to treatment with a kinase inhibitor, the method comprising:

determining the level of macrophages in a tumour sample obtained from the patient; and

comparing the level of macrophages to a predetermined threshold,

wherein a level of macrophages in the tumour sample which exceeds the threshold indicates that the cancer will respond to treatment with the kinase inhibitor.

7. The kinase inhibitor for use or a method according to any one of the preceding claims, wherein the threshold is the 75^th percentile of the distribution of macrophage levels present in tumours of the cancer.

8. The kinase inhibitor for use or a method according to claim 2 or 5, wherein the level of NK cells is the 75^th percentile of the distribution of NK cell levels present in tumours of the cancer.

9. A method of predicting the response of a cancer patient to treatment with a kinase inhibitor, the method comprising:

administering the kinase inhibitor to the patient;

determining the level of interleukin-1 b (I L-1 b) and/or interleukin-18 (IL-18) in a sample obtained from the patient; and

comparing the level of I L-1 b and/or IL-18 to a predetermined threshold,

wherein a level of I L-1 b and/or IL-18 which exceeds the threshold indicates that the patient will respond to treatment with the kinase inhibitor.

10. The kinase inhibitor for use or a method according to any one of the preceding claims, wherein the kinase inhibitor is capable of inducing pyroptosis of macrophages present in a tumour of the patient.

1 1 . The kinase inhibitor for use or a method according to any one of the preceding claims, wherein the kinase inhibitor is a type I or type II kinase inhibitor.

12. The kinase inhibitor for use or a method according to claim 1 1 , wherein the kinase inhibitor is sorafenib, or a pharmaceutically acceptable salt, hydrate, or solvate thereof.

13. The kinase inhibitor for use or method according to any one of the preceding claims, wherein the cancer is a solid cancer.

14. The kinase inhibitor for use or method according to any one of the preceding claims, wherein the cancer is hepatocellular carcinoma (HCC), renal cell carcinoma (RCC), or differentiated thyroid cancer (DTC).

15. The kinase inhibitor for use or method according to claim 14, wherein the cancer is HCC.