85 Development of a 7 plex mIF Opal panel to study solid tumor immune microenvironment

Background

Formation of tumor immune microenviroment (TIME) and interaction with tumor cells impacts response to immunotherapy. Tyramide-based multiplex immunofluorescence (mIF) on FFPE tumor samples is a powerful technology to study the concomitant expression and localization of immune biomarkers while preserving tissue architecture and spatial context in TIME. 4–1BB, an activation-induced costimulatory molecule for immune cells, is an important regulator response and being actively developed in clinical trials. The current 7 plex opal mIF study targeting 4–1BB aims to characterize the 4–1BB positive TIME, particularly 4–1BB positive B lymphocytes and their spatial relationship with myeloid and tumor cells.

Methods

The automated 7-color mIF assay (figure 1) was designed to study expression of 6 biomarkers targeting tumor or immune cells, PanCytokeratin (PanCK), 4–1BB, CD11c, CD68, CD163, CD20 and nuclear counterstain DAPI simultaneously on FFPE tissue sections of head and neck squamous cell carcinoma and non-small cell lung carcinoma. Tonsil FFPE tissue were included for sample and staining controls (figure 2). PanCK, 4–1BB, CD11c, CD68, CD163 and CD20 primary antibodies were first validated by IHC in tonsil positive control tissue (figures 3 and 4). The staining condition of the 7-color complete TSA amplification Opal mIF assay was then optimized and developed on Leica Bond auto stainer. Assay variation in HNSCC and NSCLC was determined in precision experiment (figures 5 and 6). In brief, 3 individual samples from each disease indication and 9 sequential sections from each sample were used to calculate the cell density for each biomarker and the coefficient of variations (CVs) of cell density were determined among intra/inter experimental runs.

Results

The specificity and accuracy of staining patterns of the listed biomarkers were benchmarked in control tonsil and tumor tissues by comparing the positivity and staining patterns between mIF and correspondent singleplex IHC images. Precision assessment was performed to demonstrate reproducibility of the assay. For high prevalence biomarkers, PanCK and CD20, CVs range less than 5–20% were observed in tonsil samples. CVs ranges higher than 5–20% were observed in tumor samples (figures 7–9). For 4–1BB as a less prevalent biomarker, higher CV ranges were observed in both tonsil and tumor samples (figure 9). Interestingly, 4–1BB positive B cells were found to be enriched in the germinal center (GC) of secondary lymphoid follicles in tonsil tissues, and in tertiary lymphoid structures (TLS) in tumor samples.

Conclusions

These results indicate that the Opal mIF panel is sufficient to study 4–1BB positive immune cells and their spatial relationship with myeloid and tumor cells in TIMEs.