Oops, you're using an old version of your browser so some of the features on this page may not be displaying properly.

MINIMAL Requirements: Google Chrome 24+Mozilla Firefox 20+Internet Explorer 11Opera 15–18Apple Safari 7SeaMonkey 2.15-2.23

Poster display

3820 - Biomarkers in cancer immunotherapy: analysis of clinical, histological and immunohistochemical factors associated with PD-L1 status


10 Oct 2016


Poster display


Audrey Bellesoeur


Annals of Oncology (2016) 27 (6): 15-42. 10.1093/annonc/mdw363


A. Bellesoeur, C. Massard, J. Soria

Author affiliations

  • Drug Development Department, Gustave Roussy Cancer Center, 94800 - Villejuif/FR


Abstract 3820


Inhibitors of PD-1/PD-L1 have changed the treatment paradigm in oncology. However, only a minority of patients respond to these therapies and therefore there is a critical need to identify molecular predictors that can help select patients to avoid unnecessary costs and toxicities. PD-L1 expression, detected by immunohistochemistry is currently the most explored biomarker, but its predictive value is still unclear.


Between January 2014 and July 2015, 346 patients were candidate for treatment with anti-PD-1/PD-L1 antibodies (pembrolizumab P, durvalumab D or atezolizumab A) at Gustave Roussy Cancer Campus. We reviewed in detail the files of 309 patients for whom PD-L1 expression could have been measured. Clinical, histological and IHC features were retrospectively collected for all of them in order to identify factors affecting PD-L1 status.


83 patients (26.9%) were positive for PD-L1 in IHC. PD-L1 positivity varied significantly depending on the treatment considered: P = 32.9%; D= 30.5% et A = 11.5% (p = 0.002). This suggests a relation between the companion test used for IHC and PD-L1 positivity. Univariate analysis showed that PD-L1 expression was associated with IHC test, primary tumor site and histological type. In multivariate analysis, IHC test and histological type remained significant (respectively p= 0.001 and 0.008). Atezolizumab companion test (SP142) was associated with low PD-L1 expression; while squamous cell carcinoma and urothelial carcinoma were associated with high expression. Among metastatic patients, samples from liver metastasis showed a trend for lower PD-L1 expression. There was no significant relationship with sampling method (biopsy or surgical resection), histological analysis place (private or academic), sample size, age or cellularity, or sample location (primary or metastatic site). There was no difference between samples acquired before or after some anticancer treatment.


This study identified that IHC test and histological type were associated with PD-L1 status, which need to be further investigated. The trend for lower PD-L1 expression in liver metastasis sample needs to be confirmed and compared with clinical response data.

Clinical trial identification

Legal entity responsible for the study

Gustave Roussy Cancer Campus


Gustave Roussy Cancer Campus


C. Massard: Honoraria from Genentech, Celgene, MedImmune. J-C. Soria: Consultancy fees from MSD, AstraZeneca, Roche-Genentech. All other authors have declared no conflicts of interest.

This site uses cookies. Some of these cookies are essential, while others help us improve your experience by providing insights into how the site is being used.

For more detailed information on the cookies we use, please check our Privacy Policy.

Customise settings