Digital Pathology Applications for PD-L1 Scoring in Head and Neck Squamous Cell Carcinoma: A Challenging Series

Abstract

The assessment of programmed death-ligand 1 (PD-L1) combined positive scoring (CPS) in head and neck squamous cell carcinoma (HNSCC) is challenged by pre-analytical and inter-observer variabilities. An educational program to compare the diagnostic performances between local pathologists and a board of pathologists on 11 challenging cases from different Italian pathology centers stained with PD-L1 immunohistochemistry on a digital pathology platform is reported. A laboratory-developed test (LDT) using both 22C3 (Dako) and SP263 (Ventana) clones on Dako or Ventana platforms was compared with the companion diagnostic (CDx) Dako 22C3 pharm Dx assay. A computational approach was performed to assess possible correlations between stain features and pathologists' visual assessments. Technical discordances were noted in five cases (LDT vs. CDx, 45%), due to an abnormal nuclear/cytoplasmic diaminobenzidine (DAB) stain in LDT (n = 2, 18%) and due to variation in terms of intensity, dirty background, and DAB droplets (n = 3, 27%). Interpretative discordances were noted in six cases (LDT vs. CDx, 54%). CPS remained unchanged, increased, or decreased from LDT to CDx in three (27%) cases, two (18%) cases, and one (9%) case, respectively, around relevant cutoffs (1 and 20, k = 0.63). Differences noted in DAB intensity/distribution using computational pathology partly explained the LDT vs. CDx differences in two cases (18%). Digital pathology may help in PD-L1 scoring, serving as a second opinion consultation platform in challenging cases. Computational and artificial intelligence tools will improve clinical decision-making and patient outcomes.

Keywords: PD-L1; combined positive score; digital pathology; head and neck squamous cell carcinoma.

Figure 1

Spaghetti plot showing the case-by-case modifications of the PD-L1 CPS from LDT evaluated by local pathologist, LDT evaluated by the board, and CDx evaluated by the board (a). In (b), boxplots show the distribution of PD-L1 CPS in the three evaluation sets. ID: identifier; PD-L1: programmed death ligand 1; CPS: combined positive score; LDT: laboratory-developed test; CDx: companion diagnostic.

Figure 2

Abnormal nuclear ((a), 30×) or cytoplasmic ((b), 30×) staining with original PD-L1 LDT IHC in cases n° 2 and 3, respectively. Case n° 9 was affected by major technical issues, as per board judgment, showing a less intense but sharper membrane staining with the LDT assay ((c), 30×), as compared to the CDx-based preparation ((d), 30×).

Figure 3

Examples of cases affected by interpretative variability due to technical factors. In case n° 7, sharper staining was obtained with LDT ((a), ×10), where fewer positive cells were counted, as compared to the more intense CDx-derived preparation ((b), ×10) that received a higher CPS. Case n°1 was affected by significant crush artifacts both on LDT-stained ((c), ×10) and CDx-stained ((d), ×10) tissues, complicating the single-cell localization of the positivity by pathologists.

Figure 4

Output of the computational analysis performed on QuPath. In this case (n° 10), the original IHC slide ((a), ×1) was processed for the automatic detection of tumor (green) vs. non-tumor (violet) cells ((b), ×1), allowing the subsequent comparative analysis between compartments and assays. A closer look at the tumor–stromal interface ((c), ×10) demonstrates the impact of the evaluation of immune-positive cells in this region for the elaboration of CPS, easily computable through the application of the proposed QuPath pipeline ((d), ×10).

Figure 5

The computational analysis demonstrated differences in terms of DAB staining features within the tumor and non-tumor regions between the LDT and CDx preparations, as demonstrated by the boxplots of case n° 9 (left) and n° 11 (right), confirming the possible analytical reasons for the interpretation variability described. Lab 1: original local laboratory; Lab 2: reference laboratory.