Prognostic Impact and Spatial Interplay of Immune Cells in Urothelial Cancer

BACKGROUND

Quantity and the spatial relationship of specific immune cell types can provide prognostic information in bladder cancer.

OBJECTIVE

To characterize the spatial interplay and prognostic role of different immune cell subpopulations in bladder cancer.

DESIGN, SETTING, AND PARTICIPANTS

A total of 2463 urothelial bladder carcinomas were immunostained with 21 antibodies using BLEACH&STAIN multiplex fluorescence immunohistochemistry in a tissue microarray format and analyzed using a framework of neuronal networks for an image analysis.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Spatial immune parameters were compared with histopathological parameters and overall survival data.

RESULTS AND LIMITATIONS

The identification of > 300 different immune cell subpopulations and the characterization of their spatial relationship resulted in numerous spatial interaction patterns. Thirty-nine immune parameters showed prognostic significance in univariate analyses, of which 16 were independent from pT, pN, and histological grade in muscle-invasive bladder cancer. Among all these parameters, the strongest association with prolonged overall survival was identified for intraepithelial CD8+ cytotoxic T cells (time-dependent area under receiver operating characteristic curve [AUC]: 0.70), while stromal CD8+ T cells were less relevant (AUC: 0.65). A favorable prognosis of inflamed cancers with high levels of "exhaustion markers" suggests that TIM3, PD-L1, PD-1, and CTLA-4 on immune cells do not hinder antitumoral immune response in tumors rich of tumor infiltrating immune cells.

CONCLUSIONS

The density of intraepithelial CD8+ T cells was the strongest prognostic feature in muscle-invasive bladder cancer. Given that tumor cell killing by CD8+ cytotoxic T lymphocytes through direct cell-to-cell-contacts represents the "terminal end route" of antitumor immunity, the quantity of "tumor cell adjacent CD8+ T cells" may constitute a surrogate for the efficiency of cancer recognition by the immune system that can be measured straightaway in routine pathology as the CD8 labeling index.

PATIENT SUMMARY

Quantification of intraepithelial CD8+ T cells, the strongest prognostic feature identified in muscle-invasive bladder cancer, can easily be assessed by brightfield immunohistochemistry and is therefore "ready to use" for routine pathology.

Loss of TROP2 and epithelial cell adhesion molecule expression is linked to grade progression in pTa but unrelated to disease outcome in pT2-4 urothelial bladder carcinomas

Introduction

Trophoblast cell surface antigen 2 (TROP2; EpCAM2) is a transmembrane glycoprotein which is closely related to EpCAM (EpCAM; EpCAM1). Both proteins share partial overlapping functions in epithelial development and EpCAM expression but have not been comparatively analyzed together in bladder carcinomas. TROP2 constitutes the target for the antibody-drug conjugate Sacituzumab govitecan (SG; TrodelvyTM) which has been approved for treatment of metastatic urothelial carcinoma by the United States Food and Drug administration (FDA) irrespective of its TROP2 expression status.

Methods

To evaluate the potential clinical significance of subtle differences in TROP2 and EpCAM expression in urothelial bladder cancer, both proteins were analyzed by multiplex fluorescence immunohistochemistry in combination with a deep-learning based algorithm for automated cell detection on more than 2,700 urothelial bladder carcinomas in a tissue microarray (TMA) format.

Results

The staining pattern of TROP2 and EpCAM were highly similar. For both proteins, the staining intensity gradually decreased from pTa G2 low grade (TROP2: 68.8±36.1; EpCAM: 21.5±11.7) to pTa G2 high grade (64.6±38.0; 19.3±12.2) and pTa G3 (52.1±38.7; 16.0±13.0, p<0.001 each). In pT2-4 carcinomas, the average TROP2 and EpCAM staining intensity was intermediate (61.8±40.9; 18.3±12.3). For both proteins, this was significantly lower than in pTa G2 low grade (p<0.001 each) but also higher than in pTa G3 tumors (p=0.022 for TROP2, p=0.071 for EpCAM). Within pT2-4 carcinomas, the TROP2 and EpCAM staining level was unrelated to pT, grade, UICC-category, and overall or tumor-specific patient survival. The ratio TROP2/EpCAM was unrelated to malignant phenotype and patient prognosis.

Conclusion

Our data show that TROP2 and EpCAM expression is common and highly interrelated in urothelial neoplasms. Despite of a progressive loss of TROP2/EpCAM during tumor cell dedifferentiation in pTa tumors, the lack of associations with clinicopathological parameters in pT2-4 cancer argues against a major cancer driving role of both proteins for the progression of urothelial neoplasms.

Automated Prognosis Marker Assessment in Breast Cancers Using BLEACH&STAIN Multiplexed Immunohistochemistry

Prognostic markers in routine clinical management of breast cancer are often assessed using RNA-based multi-gene panels that depend on fluctuating tumor purity. Multiplex fluorescence immunohistochemistry (mfIHC) holds the potential for an improved risk assessment. To enable automated prognosis marker detection (i.e., progesterone receptor [PR], estrogen receptor [ER], androgen receptor [AR], GATA3, TROP2, HER2, PD-L1, Ki67, TOP2A), a framework for automated breast cancer identification was developed and validated involving thirteen different artificial intelligence analysis steps and an algorithm for cell distance analysis using 11+1-marker-BLEACH&STAIN-mfIHC staining in 1404 invasive breast cancers of no special type (NST). The framework for automated breast cancer detection discriminated normal glands from malignant glands with an accuracy of 98.4%. This approach identified that five (PR, ER, AR, GATA3, PD-L1) of nine biomarkers were associated with prolonged overall survival (p ≤ 0.0095 each) and two of these (PR, AR) were found to be independent risk factors in multivariate analysis (p ≤ 0.0151 each). The combined assessment of PR-ER-AR-GATA3-PD-L1 as a five-marker prognosis score showed strong prognostic relevance (p < 0.0001) and was an independent risk factor in multivariate analysis (p = 0.0034). Automated breast cancer detection in combination with an artificial intelligence-based analysis of mfIHC enables a rapid and reliable analysis of multiple prognostic parameters. The strict limitation of the analysis to malignant cells excludes the impact of fluctuating tumor purity on assay precision.

BLEACH&STAIN 15-marker multiplexed imaging in 3098 human carcinomas reveals 6 major PD-L1-driven immune phenotypes with distinct spatial orchestration

Multiplex fluorescence immunohistochemistry (mfIHC) approaches were yet either limited to 6 markers or limited to a small tissue size that hampers translational studies on large tissue microarray cohorts. Here we have developed a BLEACH&STAIN mfIHC method that enabled the simultaneous analysis of 15 biomarkers (PD-L1, PD-1, CTLA-4, panCK, CD68, CD163, CD11c, iNOS, CD3, CD8, CD4, FOXP3, CD20, Ki67, CD31) in 3098 tumor samples from 44 different carcinoma entities within one week. To facilitate automated immune checkpoint quantification on tumor and immune cells and study its spatial interplay an artificial intelligence-based framework -incorporating 17 different deep-learning systems- was established. Unsupervised clustering showed that the three PD-L1 phenotypes (PD-L1+tumor and immune cells, PD-L1+immune cells, PD-L1 negative) were either inflamed or non-inflamed. In the inflamed PD-L1+patients, spatial analysis revealed that an elevated intratumoral M2-macrophages as well as CD11c+dendritic cell infiltration (p<0.001 each) was associated with a high CD3+CD4±CD8±FOXP3±T-cell exclusion and a high PD-1 expression on T-cells (p<0.001 each). In breast cancer, the PD-L1 fluorescence intensity on tumor cells showed a significantly higher predictive performance for overall survival (AUC: 0.72, p<0.001) compared to the commonly used percentage of PD-L1+ tumor cells (AUC: 0.54). In conclusion, our deep learning-based BLEACH&STAIN framework facilitates rapid and comprehensive assessment of more than 60 spatially orchestrated immune cell subpopulations and its prognostic relevance. Implications: The development of an easy-to-use high-throughput 15+1 multiplex fluorescence approach facilitates the in-depth understanding of the immune tumor microenvironment and enables to study the prognostic relevance of more than 130 immune cell subpopulations.

Automated Ki-67 labeling index assessment in prostate cancer using artificial intelligence and multiplex fluorescence immunohistochemistry

The Ki-67 labeling index (Ki-67 LI) is a strong prognostic marker in prostate cancer, although its analysis requires cumbersome manual quantification of Ki-67 immunostaining in 200-500 tumor cells. To enable automated Ki-67 LI assessment in routine clinical practice, a framework for automated Ki-67 LI quantification, which comprises three different artificial intelligence analysis steps and an algorithm for cell-distance analysis of multiplex fluorescence immunohistochemistry (mfIHC) staining, was developed and validated in a cohort of 12,475 prostate cancers. The prognostic impact of the Ki-67 LI was tested on a tissue microarray (TMA) containing one 0.6 mm sample per patient. A 'heterogeneity TMA' containing three to six samples from different tumor areas in each patient was used to model Ki-67 analysis of multiple different biopsies, and 30 prostate biopsies were analyzed to compare a 'classical' bright field-based Ki-67 analysis with the mfIHC-based framework. The Ki-67 LI provided strong and independent prognostic information in 11,845 analyzed prostate cancers (p < 0.001 each), and excellent agreement was found between the framework for automated Ki-67 LI assessment and the manual quantification in prostate biopsies from routine clinical practice (intraclass correlation coefficient: 0.94 [95% confidence interval: 0.87-0.97]). The analysis of the heterogeneity TMA revealed that the Ki-67 LI of the sample with the highest Gleason score (area under the curve [AUC]: 0.68) was as prognostic as the mean Ki-67 LI of all six foci (AUC: 0.71 [p = 0.24]). The combined analysis of the Ki-67 LI and Gleason score obtained on identical tissue spots showed that the Ki-67 LI added significant additional prognostic information in case of classical International Society of Urological Pathology grades (AUC: 0.82 [p = 0.002]) and quantitative Gleason score (AUC: 0.83 [p = 0.018]). The Ki-67 LI is a powerful prognostic parameter in prostate cancer that is now applicable in routine clinical practice. In the case of multiple cancer-positive biopsies, the sole automated analysis of the worst biopsy was sufficient. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Abstract P069: Semi-automated validation and quantification of CTLA-4 in 90 different Tumor entities using multiple antibodies and artificial intelligence

Introduction: CTLA-4 is an inhibitory immune checkpoint receptor and a negative regulator of anti-tumor T-cell function. This study aimed at a comparative analysis of CTLA-4+ cells between different tumor entities.

Methods: To quantify CTLA-4+ cells, 4,582 tumor samples from 90 different tumor entities as well as 608 samples of 76 different normal tissue types were analyzed by immunohistochemistry in a tissue microarray format. Two different antibody clones (MSVA-152R and CAL49) were validated and quantified using a deep learning framework for automated exclusion of unspecific immunostaining.

Results: Comparing both CTLA-4 antibodies revealed a clone dependent cytoplasmic cross-reactivity in adrenal cortical adenoma (63%) for MSVA-152R and in pheochromocytoma (67%) as well as hepatocellular carcinoma (36%) for CAL49. After automated exclusion of non-specific staining reaction (3.6%), a strong correlation was observed for the densities of CTLA-4+ lymphocytes obtained by both antibodies (r=0.87; p&lt;0.0001). The mean density of CTLA-4+ cells was 674±1482 cells/mm2 and ranged from 71±175 cells/mm2 in leiomyoma to 5916±3826 cells/mm2 in Hodgkin's lymphoma. Within epithelial tumors, the density of CTLA-4+ lymphocytes were higher in squamous cell (421±467 cells/mm2) and urothelial carcinomas (419±347 cells/mm2) than in adenocarcinomas (269±375 cells/mm2) and renal cell neoplasms (256±269 cells/mm2). A high CTLA-4+ cell density was linked to low pT category (p&lt;0.0001), absent lymph node metastases (p=0.0354), and PD-L1 expression in tumor cells or inflammatory cells (p&lt;0.0001 each). A high CTLA-4/CD3-ratio was linked to absent lymph node metastases (p=0.0295) and to PD-L1 positivity on immune cells (p&lt;0.0026).

Conclusion: Marked differences exist in the number of CTLA-4+ lymphocytes between tumors. Analyzing two independent antibodies by a deep learning framework identifies clone-specific cross-reactivity and facilitates automated quantification of target proteins such as CTLA-4.

Citation Format: David Dum, Tjark L. C. Henke, Tim Mandelkow, Elena Bady, Jonas B. Raedler, Ronald Simon, Guido Sauter, Maximilian Lennartz, Waldemar Wilczak, Eike Burandt, Stefan Steurer, Niclas C. Blessin. Semi-automated validation and quantification of CTLA-4 in 90 different Tumor entities using multiple antibodies and artificial intelligence [abstract]. In: Abstracts: AACR Virtual Special Conference: Tumor Immunology and Immunotherapy; 2021 Oct 5-6. Philadelphia (PA): AACR; Cancer Immunol Res 2022;10(1 Suppl):Abstract nr P069.

Semi-automated validation and quantification of CTLA-4 in 90 different tumor entities using multiple antibodies and artificial intelligence

CTLA-4 is an inhibitory immune checkpoint receptor and a negative regulator of anti-tumor T-cell function. This study is aimed for a comparative analysis of CTLA-4⁺ cells between different tumor entities. To quantify CTLA-4⁺ cells, 4582 tumor samples from 90 different tumor entities as well as 608 samples of 76 different normal tissue types were analyzed by immunohistochemistry in a tissue microarray format. Two different antibody clones (MSVA-152R and CAL49) were validated and quantified using a deep learning framework for automated exclusion of unspecific immunostaining. Comparing both CTLA-4 antibodies revealed a clone dependent unspecific staining pattern in adrenal cortical adenoma (63%) for MSVA-152R and in pheochromocytoma (67%) as well as hepatocellular carcinoma (36%) for CAL49. After automated exclusion of non-specific staining reaction (3.6%), a strong correlation was observed for the densities of CTLA-4⁺ lymphocytes obtained by both antibodies (r = 0.87; p < 0.0001). A high CTLA-4⁺ cell density was linked to low pT category (p < 0.0001), absent lymph node metastases (p = 0.0354), and PD-L1 expression in tumor cells or inflammatory cells (p < 0.0001 each). A high CTLA-4/CD3-ratio was linked to absent lymph node metastases (p = 0.0295) and to PD-L1 positivity on immune cells (p = 0.0026). Marked differences exist in the number of CTLA-4⁺ lymphocytes between tumors. Analyzing two independent antibodies by a deep learning framework can facilitate automated quantification of immunohistochemically analyzed target proteins such as CTLA-4.

Abstract P068: Automated cell type specific PD-L1 quantification by artificial intelligence using high throughput bleach & stain 15-marker multiplex fluorescence immunohistochemistry in human cancers

Introduction: The quantification of PD-L1 (programmed cell death ligand 1) has been used to predict patient's survival, to characterize the tumor immune microenvironment, and to predict response to immune checkpoint therapies. However, a framework to assess the PD-L1 status with a high interobserver reproducibility on tumor cells and different types of immune cells has yet to be established.

Methods: To study the impact of PD-L1 expression on the tumor immune microenvironment and patient outcome, a framework for fully automated PD-L1 quantification on tumor cells and immune cells was established and validated. Automated PD-L1 quantification was facilitated by incorporating three different deep learning steps for the analysis of more than 80 different neoplasms from more than 10'000 tumor specimens using a bleach & stain 15-marker multiplex fluorescence immunohistochemistry panel (i.e., PD-L1, PD-1, CTLA-4, panCK, CD68, CD163, CD11c, iNOS, CD3, CD8, CD4, FOXP3, CD20, Ki67, CD31). Clinicopathological parameter were available for more than 30 tumor entities and overall survival data were available for 1517 breast cancer specimens.

Results: Comparing the automated deep-learning based PD-L1 quantification with conventional brightfield PD-L1 data revealed a high concordance in tumor cells (p&lt;0.0001) as well as immune cells (p&lt;0.0001) and an accuracy of the automated PD-L1 quantification ranging from 90% to 95.2%. Across all tumor entities, the PD-L1 expression level was significantly higher in distinct macrophage/dendritic cell (DC) subsets (identified by CD68, CD163, CD11c, iNOS; p&lt;000.1) and in macrophages/DCs located in the Stroma (p&lt;0.0001) as compared to intratumoral macrophages/DC subsets. Across all different tumor entities, the PD-L1 expression was highly variable and distinct PD-L1 driven immune phenotypes were identified based on the PD-L1 intensity on both tumor and immune cells, the distance between non-exhausted T-cell subsets (i.e. PD-1 and CTLA-4 expression on CD3+CD8+ cytotoxic T-cells, CD3+CD4+ T-helper cells, CD3+CD4+FOXP3+ regulatory T-cells) and tumor cells as well as macrophage/(DC) subtypes. In breast cancer, the PD-L1 fluorescence intensity on tumor cells showed a significantly higher predictive performance for overall survival with an area under receiver operating curves (AUC) of 0.72 (p&lt;0.0001) than the percentage of PD-L1+ tumor cells (AUC: 0.54). In PD-L1 positive as well as negative breast cancers a close spatial relationship between T- cell subsets (CD3+CD4±CD8±FOXP3±PD-1±CTLA-4±) and Macrophage/DC subsets (CD68±CD163±CD11c±iNOS) was found prognostic relevant (p&lt;0.0001).

Conclusion: In conclusion, multiplex immunofluorescence PD-L1 assessment provides cutoff-free/continuous PD-L1 data which are superior to the conventional percentage of PD-L1+ tumor cells and of high prognostic relevance. The combined analysis of spatial PD-L1/PD-1 data and more than 20 different immune cell subtypes of the immune tumor microenvironment revealed distinct PD-L1 immune phenotypes.

Citation Format: Niclas C. Blessin, Elena Bady, Tim Mandelkow, Cheng Yang, Jonas B. Raedler, Ronald Simon, Christoph Fraune, Maximilian Lennartz, Sarah Minner, Eike Burandt, Doris Höflmayer, Guido Sauter, Katharina Möller, Sören A. Weidemann. Automated cell type specific PD-L1 quantification by artificial intelligence using high throughput bleach & stain 15-marker multiplex fluorescence immunohistochemistry in human cancers [abstract]. In: Abstracts: AACR Virtual Special Conference: Tumor Immunology and Immunotherapy; 2021 Oct 5-6. Philadelphia (PA): AACR; Cancer Immunol Res 2022;10(1 Suppl):Abstract nr P068.

Prognostic role of proliferating CD8+ cytotoxic Tcells in human cancers

PURPOSE

Expansion of CD8⁺ cytotoxic Tlymphocytes is a prerequisite for anti-cancer immune activity and has gained interest in the era of immune checkpoint therapy.

METHODS

To understand the CD8⁺ T cell dynamics in the tumor microenvironment, we used multiplex fluorescence immunohistochemistry to quantitate CD8⁺ proliferation (Ki67 co-expression) in tissue microarrays from 1107 colorectal, 642 renal cell, 1066 breast, 375 ovarian, 451 pancreatic and 347 gastric cancer samples.

RESULTS

The density and the percentage of proliferating (Ki67⁺) CD8⁺ T cells were both highly variable between tumor types as well as between patients with the same tumor type. Elevated density and percentage of proliferating CD8⁺ cytotoxic T cells were significantly associated with favorable tumor parameters such as low tumor stage, negative nodal stage (p ≤ 0.0041 each), prolonged overall survival (p ≤ 0.0028 each) and an inflamed immune phenotype (p = 0.0025) in colorectal cancer and, in contrast, linked to high tumor stage, advanced ISUP/Fuhrman/Thoenes grading (each p ≤ 0.003), shorter overall survival (p ≤ 0.0330 each) and an immune inflamed phenotype (p = 0.0094) in renal cell cancer. In breast, ovarian, pancreatic and gastric cancer the role of (Ki67⁺)CD8⁺ Tcells was not linked to clinicopathological data.

CONCLUSION

Our data demonstrate a tumor type dependent prognostic impact of proliferating (Ki67⁺)CD8⁺ Tcells and an inverse impact in colorectal and renal cell cancer.