Abstract LB-248: RNA-sequencing of tumor-educated platelets enables nivolumab immunotherapy response prediction

• Published in: Cancer research (Chicago, Ill.) Vol. 77; no. 13_Supplement; p. LB-248
• Main Authors: Muller, Mirte, Best, Myron, Sol, Nik, Niemeijer, Anna-Larissa N., Vancura, Adrienne, Schouten, Robert D., Hiltermann, Jeroen J.N., Veld, Sjors G.J.G. In 't, Broek, Daan van den, Noort, Vincent van der, Langen, Adrianus J. de, Schuuring, Ed M.D., Wurdinger, Thomas, Heuvel, Michel M. van den
• Format: Journal Article
• Language: English
• Published: 01.07.2017
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2017-LB-248

Abstract I: Studies have shown the activity of anti-PD(L)-1 therapies in patients with advanced non-small-cell lung cancer (NSCLC), however response rates are rather low (~20%). Therefore, there is an urgent need to identify biomarkers that predict patient outcome to immunotherapy (IT). Previously we have shown that RNA signatures of tumor-educated platelets (TEPs) may have predictive value for tumor type-specific diagnostics. Platelets are intimately involved in immune responses. We hypothesized that TEP RNA profiles have predictive value for IT response. M: We collected baseline platelet pellets, isolated from whole blood by differential centrifugation, from 389 stage IV NSCLC patients treated with Nivolumab (table 1). Tumor response was evaluated at 3 and 6 months and reported according to RECIST 1.1. Platelet pellets were subjected to total RNA isolation, SMARTer cDNA amplification, and libraries were sequenced on the HiSeq platform. Raw data (~20 M reads per sample) was mapped to the human genome, intron-spanning spliced RNA reads were selected for analysis. Gene panels were calculated by ANOVA statistics. R: Until now 64 samples were sequenced, 30 of patients with clinical benefit (PR or SD at six months; CB) and 34 of patients with no clinical benefit (PD; no CB). 40 randomly selected samples (20 CB, 20 no CB) were used for training of the support vector machine (SVM)-based therapy response classification algorithm. 24 samples were used for independent evaluation of the classifier. Hierarchical clustering of genes with p<0.05 among these groups identified significant separation between CB and no CB (p<0.001, Fisher’s exact test). The SVM nivolumab response prediction algorithm, predicted the remaining 24 samples with an accuracy of more than 85%, as opposed to at random classification (p<0.01). D: TEP RNA profiles potentially enable liquid biopsy-based response prediction to IT. Large-scale validation is ongoing and up to date results of the NSCLC cohort will be presented. Table 1.Patient CharacteristicsTotal CohortNumber of patients389Start of treatmentAugust 2015-november 2016Date lock1th of November, 2016Gender (Male)56%Age (med, range)64.5 years (29-83)Treatment lines1st line1%2nd line72%>2nd line27%DiagnosisNon-Squamous68%Squamous26%Other6%ResponseClinical benefit27%Progressive54%Unknown19%PD-L1 expressionResults expected in april Citation Format: Mirte Muller, Myron Best, Nik Sol, Anna-Larissa N. Niemeijer, Adrienne Vancura, Robert D. Schouten, Jeroen J.N. Hiltermann, Sjors G.J.G. In 't Veld, Daan van den Broek, Vincent van der Noort, Adrianus J. de Langen, Ed M.D. Schuuring, Thomas Wurdinger, Michel M. van den Heuvel. RNA-sequencing of tumor-educated platelets enables nivolumab immunotherapy response prediction [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr LB-248. doi:10.1158/1538-7445.AM2017-LB-248