Interleukin-17-positive mast cells influence outcomes from BCG for patients with CIS: Data from a comprehensive characterisation of the immune microenvironment of urothelial bladder cancer

Abstract

The tumour immune microenvironment is considered to influence cancer behaviour and outcome. Using a panel of markers for innate and adaptive immune cells we set out to characterise and understand the bladder tumour microenvironment of 114 patients from a prospective multicentre cohort of newly-diagnosed bladder cancer patients, followed-up for 4.33±1.71 years. We found IL-17-positive cells were significantly increased in primary and concomitant carcinoma in situ (CIS), p<0.0001, a highly malignant lesion which is the most significant single risk factor for disease progression. Further characterisation of the tumour immunophenotype identified IL-17+ cells as predominantly mast cells rather than T-cells, in contrast to most other tumour types. Expression of the IL-17-receptor in bladder tumours, and functional effects and gene expression changes induced by IL-17 in bladder tumour cells in vitro suggest a role in tumour behaviour. Finally, we assessed the effects of IL-17 in the context of patient outcome, following intravesical BCG immunotherapy which is the standard of care; higher numbers of IL-17+ cells were associated with improved event-free survival (p = 0.0449, HR 0.2918, 95% CI 0.08762-0.9721) in patients with primary and concomitant CIS (n = 41), we propose a model of IL-17+ Mast cells mechanism of action. Thus, in the context of bladder CIS, IL-17+ mast cells predict favourable outcome following BCG immunotherapy indicative of a novel mechanism of BCG immunotherapy in UBC and could form the basis of a stratified approach to treatment.

Fig 1. Immunohistochemical analysis of the bladder cancer immune microenvironment.

Representative results from bladder cancer FFPE biopsy sections stained using primary antibodies specific for the indicated immune cell markers. Positive cells are stained with DAB (brown) and all slides are counterstained with haematoxylin (blue). Negative control images (inset) were obtained by substituting an isotype control antibody for the primary antibody. A: The bladder cancer microenvironment contains CD3+ T cells, FoxP3+ cells, CD68+ macrophages and CD15+ granulocytes. B: Representative images from two bladder cancer biopsies with high (left) or low (right) numbers of IL-17 positive cells present. Note the different distributions of IL-17+ cells and CD3+ cells shown in A.

Fig 2. Association between IL-17 positive cells and bladder cancer grade and stage.

Upper panels: The number of IL-17 positive cells present in different bladder cancer biopsies are displayed according to (A) grade at time of diagnosis (G1, n = 10; G2, n = 9; G3 n = 95) or (B) stage at time of diagnosis (pTa, n = 17; pT1, n = 10; T2+, n = 8; CIS, n = 83). Horizontal bars indicate the mean number of cells present in all biopsies within each subgroup. Lower panels: Results of quantitative rtPCR measuring RNA transcripts present in whole snap frozen biopsies diagnosed as stage pTa/pT1(n = 5) or pTa/pT1 with concomitant CIS (n = 3). The ΔCT values of mRNA transcripts of (C) IL-17A, (D) IL-17-F, (E) IL-6 or (F) IL-23 are shown relative to GAPDH (ΔCT = CT_Experimental-CT_GAPDH). Asterisks indicate significance calculated by one way ANOVA with Dunn’s multiple comparison test (B) or unpaired T-test (C-F): *p<0.05, **p<0.01, ****p<0.0001, n/s = not significant.

Fig 3. IL-17 receptor expression and function in bladder cancer.

A: Representative result obtained for a bladder cancer biopsy, in this case CIS, stained with a primary antibody specific for the IL-17 receptor (brown staining represents receptor expression). Inset image shows result obtained using an isotype control primary antibody. B: Histograms showing levels of IL-17 receptor on surface of the urothelial cell lines EJ, 5637 or HB-CLS-2 measured using flow cytometry. Red histogram, anti-IL-17-receptor antibody; open histogram, isotype control antibody. C & D: IL-17 treatment increases production of IL-6 and IL-8 by all three urothelial cell lines tested. Cytokine levels were measured by ELISA. Error bars represent the standard deviation of the mean, which was calculated from three (IL-6) or four (IL-8) independent experiments. Asterisks indicate significance calculated by repeated measure ANOVA with Bonferroni multiple comparison test: *p<0.05, **p<0.01, ***p<0.001.. E: Growth of EJ, 5637 and HB-CLS-2 cells after five days incubation in the indicated amounts of IL-17. Cell growth was measured by WST assay. Four independent experiments were performed and the mean change in WST-1 absorbance relative to mock treated cells is shown. Error bars indicate standard deviation. F: Migration of the above cells in a 6 hour wound healing assay in the absence or presence of the indicated concentrations of IL-17. Cells were grown in IL-17 for 48 hours before wounding. The breast cancer cell line MBA-MD231, which is more invasive in response to IL-17, was included as a positive control. Three independent experiments were performed and the mean wound closure is shown, error bars represent standard deviation. In both panels: *p<0.05 calculated by a repeated measure ANOVA with Bonferroni multiple comparison test. G: Venn diagrams showing gene expression changes induced by IL-17 treatment of EJ cells (light grey circles) or 5637 cells (dark grey circles) measured in a gene array. The number of genes that undergo similar changes in expression following IL-17 treatment are shown in the intersect, and are also listed below each Venn diagram.

Fig 4. Identification of the IL-17 producing cells present in CIS.

A: Representative results of co-staining for IL-17 and cell lineage markers (combined IF and bright field images. Red fluorescence indicates IL-17 and brown staining indicates CD3 (left image), CD15 (middle) or CD68 (right image). Note that CD15 and IL-17 co-localise as indicated by arrows. B: The region indicated by dashed lines in the IL-17/CD15 co-staining image above is magnified and shown in bright field (left), IF (middle) or combined IF/bright field (right). Arrows indicate the location of co-staining cells visualised in the combined IF/bright field image. C: Characterising the CD15-positive cells present in CIS. Representative staining results from a single CIS biopsy containing a large number of CD15 positive granulocytes (left panel). This biopsy also contains MCs that stain positive for MC tryptase (MCT). No ELANE-positive neutrophils were present (right panel). In each image positive cells are stained brown. The total number of biopsies stained for CD15, ELANE or MCT were 15, 18 and 12 respectively. The inset images show sections stained using appropriate isotype control primary antibodies. D: Co-localisation of MCT and IL-17 staining in CIS biopsies. Upper image shows the result obtained for a CIS biopsy co-stained with anti-IL-17 and anti-MCT antibodies (red and dark brown staining respectively). Note how the weaker IL-17 signal is obscured by the dark brown staining of the MCT-positive cells. Lower image shows a serial section from the same tumour specimen stained for IL-17 alone (red) for comparison. The result is representative of eight CIS biopsies analysed. E: Result of co-staining a CIS biopsy for IL-17 (red) and reduced intensity MCT (brown). Arrows indicate MCT-positive MCs that co-stain for IL-17 (magnification x1000). Image is representative of the four CIS biopsies stained using the modified protocol.

Fig 5. The effect of tumoral IL-17 positive cells upon patient outcome.

A: Patients who had NMIBCs with concomitant CIS were divided into IL-17hi (n = 14) and IL-17low (n = 52) groups based on the upper quartile of IL-17 positive cell counts in this group of patients (374.5 IL-17 positive cells per 10 high power fields). B: Patients who received BCG immunotherapy were divided into IL-17hi (n = 6) and IL-17lo (n = 36) groups, using the same criteria as before. Significance was calculated using the Log-rank (Mantel-Cox) test. Dotted line indicates 95% CI.

Fig 6. Proposed mechanism for the role of IL-17-positive cells in the efficacy of BCG immunotherapy.