Abstract
Abstract 5079
Epstein-Barr virus (EBV) is associated with a number of human malignancies, including a variety of lymphomas, post-transplant lymphoproliferative disorder, and nasopharyngeal carcinoma. EBV viral load can be measured in peripheral blood by commercially available PCR assays. We seek to establish the correlation of EBV viral load with clinical disease activity in these malignancies.
We prospectively enrolled patients with radiographically measurable malignancies of any of the above-mentioned types, with preference for enrollment of subtypes known to be strongly EBV-associated. A total of 65 patients signed consent, and 57 (67% newly diagnosed, 28% pretreated, 5% unknown treatment) were enrolled after fulfillment of inclusion and exclusion criteria. A 6-primer quantitative PCR panel was used, testing for EBER, LMP and EBNA both in patient plasma (P) and in whole blood cells (WB). Patients with detectable EBV PCR at baseline with any of the 6 assays were followed longitudinally as they underwent therapy. Disease status (stable/response/progression/CR) was determined by standard clinical/radiological assessments, and was correlated with EBV viral load.
Of 57 enrolled patients, 34 (60%) had detectable EBV viral load by at least one of the 6 assays of the PCR panel. EBNA-WB, EBER-WB, and EBNA-P were the EBV assays most likely to be positive at baseline (97%, 79% and 71% respectively), with median viral loads of 650 (range 100–131,000), 200 (100-55,800) and 450 (100-134,000) copies/ml respectively. Prevalence of detectable EBV viral load at baseline varied between different malignancies (‘n' for each in brackets) – 100% for PTLD(2) and AITL(6), 75% for nasopharyngeal carcinoma(12), 63% for Hodgkin disease(8), 60% for CTCL(5), 50% for DLBCL(4), Burkitt's lymphoma(2), anaplastic TCL(4), NK/T cell lymphoma(4), 33% for PTCL(3), and 0% for follicular lymphoma(3).
LMP immunohistochemical staining and EBER in-situ hybridization were performed (based on tissue availability) on samples from 63% of enrolled patients. EBV was detected in 64% of tested tissue samples. Of tissue EBV positive patients, 87% had detectable blood EBV PCR. Notably, even in tissue EBV negative patients, 38% had detectable blood PCR.
Longitudinally measured viral loads were available for 27 of 34 EBV PCR positive patients, with a median of 3 tests (range 2–7) over a period of 4 months (range 0.3–14.9 months). Reduction of plasma EBV viral load to 200 copies/ml or less in the course of treatment had the strongest correlation with achievement of clinical CR (Spearman's rho 0.78, p<0.001) with the following test characteristics: PPV 87%, NPV 100%, area under ROC curve 93%.
More than half of our patients with a variety of lymphomas, PTLD and nasopharyngeal carcinoma had detectable peripheral blood EBV PCR. Most patients with tissue EBV positivity had a detectable blood viral load, but, surprisingly, more than a third of tissue EBV negative patients also had a detectable viral load, arguing for a potential role for routine blood EBV PCR testing even in the absence of tissue EBV positivity.
Quantitative EBV PCR had a statistically significant correlation with clinical disease activity, and achievement of plasma EBV viral load of 200 copies/ml or less had an excellent predictive value for achievement of clinical CR. We recommend further trials exploring the role of EBV PCR in guiding duration and number of cycles of chemotherapy, and potentially using EBV PCR as a modality for early detection of relapse.
Mato:Celgene: Speakers Bureau; Milennium: Speakers Bureau. Arnoldi:ViraCor: Employment.
Author notes
Asterisk with author names denotes non-ASH members.
