IADR Abstract Archives
Detection of PD-L1 expression on HNSCC cells
Objectives: Objective: The Programmed Death (PD) pathway, involving receptor PD-1 on T-cells and ligand PD-L1 on tumor cells is an important immunotherapeutic checkpoint to inhibit in the control of head and neck squamous cell carcinoma (HNSCC).
Methods: A computational simulation model was developed to predict immunosuppressive biomarker profiles, including PD-L1. HNSCC (UM-SCC) cells grown in culture were used to confirm the responses. 1.0 × 106 cells of SCC4, SCC15, SCC25, SCC19, SCC84, SCC92, and SCC99 were incubated in polypropylene tubes and culture plates for 24 hours. SCC cells in tubes were processed for microtomy, and treated with antibodies to PD-L1 in immunohistochemistry (IHC) procedures. SCC cells adhered to culture plates were lysed and concentrations of PD-L1 in cell lysates were determined (ELISA, American Research Products, Inc., Waltham, MA). An analogous two-way fixed effect ANOVA was fit to the log-transformed concentrations of PD-L1 (JMP10, SAS, Cary, NC) and pairwise group comparisons were conducted at a 0.05 level using the method of Tukey’s Honest Significant Differences.
Results: Overall, there was 85% correlation between predictive vs. experimental biomarker trends for SCC4, SCC15, and SCC25 when comparing 2 cell lines at a time. PD-L1 was detected by IHC and the staining of cells varied from 4.0-73.0% depending upon the cell-line (average staining, 30.7%). Cell lysates contained significantly different (p<0.05) concentrations of PD-L1 and ranged from 92.65 pg/ml (+38.13 pg/ml, SEM) for SCC19 to 387.59 pg/ml (+38.13 pg/ml, SEM) for SCC25.
Conclusions: Expression of PD-L1 by HNSCC cell lines can be determined based on their profile of genomic aberrations using a computational approach. Predicting PD-L1 responses of patient HNSCC cells will be an important step in identifying the role of the PD pathway in the response of patients to immunotherapy.
Methods: A computational simulation model was developed to predict immunosuppressive biomarker profiles, including PD-L1. HNSCC (UM-SCC) cells grown in culture were used to confirm the responses. 1.0 × 106 cells of SCC4, SCC15, SCC25, SCC19, SCC84, SCC92, and SCC99 were incubated in polypropylene tubes and culture plates for 24 hours. SCC cells in tubes were processed for microtomy, and treated with antibodies to PD-L1 in immunohistochemistry (IHC) procedures. SCC cells adhered to culture plates were lysed and concentrations of PD-L1 in cell lysates were determined (ELISA, American Research Products, Inc., Waltham, MA). An analogous two-way fixed effect ANOVA was fit to the log-transformed concentrations of PD-L1 (JMP10, SAS, Cary, NC) and pairwise group comparisons were conducted at a 0.05 level using the method of Tukey’s Honest Significant Differences.
Results: Overall, there was 85% correlation between predictive vs. experimental biomarker trends for SCC4, SCC15, and SCC25 when comparing 2 cell lines at a time. PD-L1 was detected by IHC and the staining of cells varied from 4.0-73.0% depending upon the cell-line (average staining, 30.7%). Cell lysates contained significantly different (p<0.05) concentrations of PD-L1 and ranged from 92.65 pg/ml (+38.13 pg/ml, SEM) for SCC19 to 387.59 pg/ml (+38.13 pg/ml, SEM) for SCC25.
Conclusions: Expression of PD-L1 by HNSCC cell lines can be determined based on their profile of genomic aberrations using a computational approach. Predicting PD-L1 responses of patient HNSCC cells will be an important step in identifying the role of the PD pathway in the response of patients to immunotherapy.