Immune-checkpoint gene expression and BCG response in non-muscle invasive bladder cancer

Abstract

Methods: One-hundred-six patients diagnosed with non-muscle invasive bladder cancer and treated with intravesical BCG were included and divided into two groups, BCG-responsive (n = 47) and -unresponsive (n = 59). Immunohistochemistry was used to evaluate PD-L1 expression and MSI was assessed by a commercial multiplex PCR kit. The mRNA expression profile of 15 immune checkpoints was performed using the nCounter technology. For in silico validation, two distinct cohorts sourced from the Gene Expression Omnibus (GEO) database were used.

Results: Among the 106 patients, only one (<1 %) exhibited MSI instability. PD-L1 expression was present in 9.4 % of cases, and no association was found with BCG-responsive status. We found low gene expression of canonic actionable immune checkpoints PDCD1 (PD-1), CD274 (PD-L1), and CTLA4, while high expression was observed for CD276 (B7-H3), CD47, TNFRSF14, IDO1 and PVR (CD155) genes. High IDO1 expression levels was associated with worst overall survival. The PDCD1, CTLA4 and TNFRSF14 expression levels were associated with BCG responsiveness, whereas TIGIT and CD276 were associated with unresponsiveness. Finally, CD276 was validated in silico cohorts.

Conclusion: In NMIBC, MSI is rare and PD-L1 expression is present in a small subset of cases. Expression levels of PDCD1, CTLA4, TNFRSF14, TIGIT and CD276 could constitute predictive biomarkers of BCG responsiveness.

Keywords: BCG vaccine (2); Gene expression profiling(3); Immunity active(4); Urinary bladder neoplasms(1).

Fig. 1

Illustrative cases with positive PD-L1 immunoexpression. A and B are the same case. A) Large amount of tumor cells exhibiting membrane positivity (TPS ≥ 1 % - positive), 400X objective. B) Mononuclear cells, mainly lymphocytes (arrows), with membrane positivity, inflammatory cell score (IC Score ≥ 1 % - positive), 400X objective. C and D) Focal tumor cells revealing membrane positive expression (arrows) (both TPS ≥ 1 %), 400X objective. All cases are CPS ≥ 1 (positive).

Fig. 2

mRNA expression heatmap of 15 immune checkpoint-related genes in the NMIBC cohort. The heatmap illustrates the mean normalized mRNA counts of the examined genes, using a color scale. Blue indicates counts below the background threshold, white signifies low counts (below 100), and red indicates higher counts. The heatmap was generated using GraphPad Prism 8 software.

Fig. 3

Association between the mRNA levels of CD274 (PD-L1) and immunohistochemistry expression of PD-L1 according to the Combined Positive Score (CPS). The x-axis represents the expression of PD-L1 according to CPS and the y-axis represents the mRNA expression of the CD274 gene assessed using NanoString. The significance level is p < 0.05.

Fig. 4

mRNA levels for PDCD1 (PD-1), CTLA4, CD276 (B7-H3), TNFRSF14 and TIGIT across BCG-unresponsive and responsive groups of 106 NMIBC samples using the nCounter technology. The significance level is 0.05. * < 0.05; ** < 0.01.

Fig. 5

Expression of immune regulatory markers in BCG-responsive and -unresponsive patients. A) Expression of FOXP3 (marker of regulatory T cells) in tumor tissue samples from patients who responded (in orange) and did not respond (in gray) to BCG therapy, for each cohort. B) Expression of ITGAM (marker of myeloid-derived suppressor cells) in tumor tissue samples from patients who responded (in orange) and did not respond (in gray) to BCG therapy, for each cohort. The significance level is 0.05. * < 0.05; ** < 0.01; *** < 0.001; ns: not significant.

Fig. 6

CD276 expression levels comparing BCG-responsive and unresponsive patients from the discovery and validation cohorts. The significance level is 0.05. * < 0.05; ** < 0.01.