Related Experiment Video
Updated: Jan 2, 2026

Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
Prognostic Immune Cell Profiling of Malignant Pleural Effusion Patients by Computerized Immunohistochemical and
Chengguang Wu1, Fabian Mairinger2, Ruben Casanova1
1Institute of Pathology and Molecular Pathology, University Hospital Zurich, CH-8091 Zurich, Switzerland.
Abstract:
Malignant pleural effusion (MPE) is a severe condition of advanced tumors without effective therapy. We used digitalized immunohistochemical and transcriptional approaches to investigate the prognostic influence of immune cells and expression variance of associated immunomodulatory molecules in MPE. Cytology tissue microarrays were constructed from MPE cell blocks of 155 patients with five tumor entities. Immune cells lineage markers were quantified by computational cytopathology on immunohistochemistry. mRNA expression analysis of nine lineage markers and 17 immunomodulators was performed by NanoString. Immunohistochemically quantified high B cells to leukocytes ratio (hazard ratio (HR) = 0.70, p-value = 0.043) and low neutrophils to leukocytes ratio (HR = 1.78, p-value = 0.003) were favorable prognosticators for overall survival independent of tumor entity. Correspondingly, patients with high B cells but low neutrophils gene expression signature showed longer median overall survival of 500 days (HR = 2.29, p-value = 0.009). Regarding targetable molecule expressions, lung adenocarcinomas were characterized by high PD-L1, but mesothelioma by high LAG-3. Ovarian carcinoma was least immunogenic. Independent of tumor entity, the condition of the immune system in MPE liquids is able to provide additional prognostic cytologic information. Combined analysis of lineage specific markers and related immunomodulators may direct immune-based therapeutic decisions.

