Abstract 4483
Background
Tumors may suppress host defenses by activating immune checkpoints (eg, the programed cell death [PD-1/PD-L1] pathway). Colocalization (CL) is a requirement for PD-1/PD-L1 interaction. PD-1/PD-L1 CL in tissue sections, as determined by immunohistochemistry (IHC), may be an indicator of PD-L1/PD-1 pathway activity.
Methods
We assessed CL of PD-L1 and PD-1 in situ by applying a novel duplex bright-field IHC technique on 49 formalin-fixed, paraffin-embedded HCC samples using digital image analysis (DIA; Definiens Tissue Studio®) to determine the percentage of single PD-1+ and PD-L1+ cells, PD-L1/PD-1 double-labeled cells, and PD-1+ cells adjacent to ≥ 1 PD-L1+ cells.
Results
All cases showed typical HCC morphology (low- to high-grade trabecular [4/49], pseudoglandular [1/49], solid [40/49], clear cell [2/49], or desmoplastic [2/49]). PD-L1 was largely observed in immune cell infiltrates. On average, 2.6% ± 3.6% (median, 1.5%) of the cells (immune + tumor) within the tumor area were PD-1+, and 4.3% ± 5.5% (median, 1.9%) were PD-L1+. There was considerable variation among samples in the number of PD-1+ (range, 0.05%-21.2%) and PD-L1+ (range, 0.2%-30.3%) cells. In 18/49 cases (37%), the number of PD-1+ cells exceeded the number of PD-L1+ cells; in 31/49 cases (63%), the number of PD-L1+ cells exceeded the number of PD-1+ cells. PD-1/PD-L1 double-stained cells were present in 31/49 cases (63%), and 1.6% ± 4.1% (median, 0.13%) of the cells were double labeled, with considerable intersample variation (range, 0.5%-22.9%). Finally, 10.5% ± 8.03% (median, 9.4%) of PD-1+ cells were in the immediate vicinity of a PD-L1+ cell (range, 1.1%-43.3%).
Conclusions
By combining a novel duplex bright-field IHC technique with DIA, we quantitated the number/distribution of PD-1+ and PD-L1+ cells in HCC. Variation in the numbers of PD-1+ and PD-L1+ cells, and PD-1+ cells with ≥1 PD-L1+ adjacent cells, in HCC was seen. Future studies can use these techniques to explore the predictive potential of PD-L1/PD-1 expression in patients who are being considered for immunotherapy. Our proof of concept results suggest that the methods may also be applied for other tumors.
Clinical trial identification
Legal entity responsible for the study
Merck KGaA, Darmstadt, Germany.
Funding
Merck KGaA, Darmstadt, Germany.
Editorial Acknowledgement
Medical writing support was provided by ClinicalThinking, and was funded by Merck KGaA, Darmstadt, Germany.
Disclosure
T. Mrowiec, F. Wilm, E. Frick-Krieger, C. Ihling: Employment: Merck KGaA. M. Silva: Employment: Merck KGaA; Equity ownership: Merck KGaA. L. Terracciano: Consultancy: Merck KGaA. I. Dussault: Employment: EMD Serono.