Prevalence of the HPV, EBV, and TTV viral RNA in the plasma of patients with solid and hematologic neoplasms and the detection of a specific immune signature.

Abstract

2583 Background: Epstein−Barr virus (EBV) and human papillomavirus (HPV) are considered human oncoviruses. In contrast, the torque teno virus (TTV) is not associated with any disease but its detection in circulation is associated with the status of the immune system. In this study, we examine the prevalence of active EBV, HPV and TTV viral RNA in patients treated for solid tumors or hematologic neoplasms. In addition, we compared differential expression of selected immune and inflammatory biomarkers between Virus positive (V+) and Virus negative (V-) cases using peripheral blood cell-free RNA (cfRNA). Methods: cfRNA was extracted from the peripheral blood of 581patients with a diagnosis of hematologic neoplasms and 558 patients with solid tumor. cfRNA was sequenced by NGS using a targeted RNA panel of 1600 genes and the viral RNA of TTV, EBV and HPV. Two thirds of the samples were used for training and one third for testing machine learning (ML) system (Bayesian/Random Forest) and exploring the presence of specific inflammatory profiles distinguishing V+ from V- patients. Results: RNA testing was selected to ensure that only active and proliferating viruses were detected. We detected TTV in 52/1139 (4.6%), EBV in 251/1139 (22%), and HPV in 68/1139 (6.0%). TTV with EBV codetection was observed in 11 samples (1%), and with HPV in 4 patients (0.4%). Co-detection of EBV with HPV was observed in 13 patients (1.1%). Using 90 biomarkers in ML algorithm can reliably distinguish V+ from V- with AUC of 0.725 (CI: 0.658-0.791) in the testing set. Significantly higher levels of B-cell markers are noted in V+ patients. PD-L1 mRNA was significantly (P <0.001) higher in V+ patients, which suggests that these patients may be more responsive to checkpoint immune therapy. CD70 is also detected at high level in V+ patients (P<0.0001). Upon comparing between the V+ groups (TTV, EBV, and HPV), there was no statistical difference between the three groups after adjusting for multiple testing. However, some difference in cytokine levels was noted between TTV-positive patients and HPV-positive patients. CD36, IFNA2 and IL17A were higher in TTV-positive cases as compared with HPV-positive cases (P-value 0.0003, 0.005 and 0.004, respectively). Conclusions: Globally, detectable active viruses in plasma of patients with cancer is relatively high (29%) and this detection is associated with a specific immune/inflammatory “activation” signature characterized by transcriptomic upregulation of PD-L1 and CD70 and increase in B-cells. There is no specific signature that distinguishes between the V+ subgroups [TTV vs EBV vs HPV]. However, transcriptionally, CD36, IFNA2 and IL17A upregulation distinguished TTV+ and HPV+ cases, a phenomenon that may indicate HPV ability to initiate an immunosuppressive tumor microenvironment.