Digital Pathology and PD-L1 Testing in Non Small Cell Lung Cancer: A Workshop Record.

Fabio Pagni,Elena V Vigliar,Paolo Graziano,Umberto Malapelle,Elena Guerini Rocco,Filippo Fraggetta,Fiamma Buttitta,Paolo Graziano,Marta Jaconi,Gabriella Fontanini,Pasquale Pisapia,Nicola Fusco,Massimo Barberis,Nicola Fusco,Massimo Barberis,Giancarlo Troncone,Elena Guerini Rocco,Claudio Doglioni,Antonio Marchetti

doi:10.3390/cancers12071800

Abstract

A meeting among expert pathologists was held in 2019 in Rome to verify the results of the previous harmonization efforts on the PD-L1 immunohistochemical testing by scoring a representative series of non-small cell lung cancer (NSCLC) digital slides. The current paper shows the results of this digital experimental meeting and the expertise achieved by the community of Italian pathologists. PD-L1 protein expression was determined using tumor proportion score (TPS), i.e., the percentage of viable tumor cells showing partial or complete membrane staining at any intensity. The gold standard was defined as the final PD-L1 score formulated by a panel of seven lung committed pathologists. PD-L1 status was clustered in three categories, namely negative (TPS < 1), low (TPS 1–49%), and high (TPS ≥ 50%). In 23 cases (71.9%) PD-L1 staining was performed using the companion diagnostic 22C3 pharmDx kit on Dako Autostainer, while in nine (28.1%) cases it was performed using the SP263 Ventana kit on BenchMark platform. A complete PD-L1 scoring agreement between the panel of experts and the participants was reached in 57.1% of cases, whereas a minor disagreement in 16.1% of cases was recorded. Italian pathologists performed best in strong positive cases (i.e., tumor proportion score TPS > 50%), whereas only 10.8% of disagreement with the gold standard was observed, and 55.6% regarded a single challenging case. The worst performance was achieved in the negative cases, with 32.0% disagreement. A significant difference resulted from the analysis of the data separated by the different clones used: 22.3% and 38.1% disagreement (p = 0.01) was found in the group of cases analyzed by 22C3 and SP263 antibody clones, respectively. In conclusion, this workshop record proposed the application of a digital pathology platform to share controversial cases in educational meetings as an alternative possibility for improving the interpretation and reporting of specific histological tools. Due to the crucial role of PD-L1 TPS for the selection of patients for immunotherapy, the identification of unconventional approaches as virtual slides to focus experiences and give more detailed practical verifications of the standard quality reached may be a considerable option.

Highlights

During the last two decades, a significant revolution in pathology laboratories has been represented by the implementation of whole slide imaging and digital scanners [1]
The original programmed death-ligand 1 (PD-L1) protein expression was determined using tumor proportion score (TPS), i.e., the percentage of viable tumor cells showing partial or complete membrane staining at any intensity
Efforts have to be devoted to validating the use of digital slides, as reported by The College of American Pathologists (CAP) and the Digital Pathology Association (DPA) guidelines [17,18]

Summary

Introduction

During the last two decades, a significant revolution in pathology laboratories has been represented by the implementation of whole slide imaging and digital scanners [1]. Digital pathology may play a significant role as a didactic, diagnostic and research tool. Digital slides are more efficient to adequately replicate the microscope experience than static images [2]. Another advantage in this setting is represented by the possibility to highlight and analyze interesting areas during didactic sessions [2]. The evaluation of the diagnostic performances of a single tissue biomarker like programmed death-ligand 1 (PD-L1) seems to be a good opportunity to test in parallel the application of digital pathology as a routine tool in educational programs Despite a more realistic experience can be obtained by adopting multi-headed microscopes, a limitation in this approach is related by the low number of viewing head extensions (5 to 14) [2] while digital slides may allow, by using dedicated monitor, an adequate experience to a higher number of attendants [3].

Methods

Discussion

Conclusion