Post-RT plasma EBV DNA sequencing improves NPC risk stratification

Targeted sequencing of plasma Epstein-Barr virus (EBV) DNA at 6–8 weeks after radiotherapy (RT) enhances risk stratification of nasopharyngeal carcinoma (NPC) compared with conventional quantitative PCR (qPCR), researchers at the Chinese University of Hong Kong (CUHK) have shown.

Notably, among patients with undetectable plasma EBV DNA on qPCR, sequencing-based analysis further defined a subgroup with superior 5-year survival outcomes. [Ann Oncol 2022;33:794-803]

The study included 769 patients with stage IIB–IVB locoregionally advanced NPC who had completed curative-intent RT or chemoradiotherapy in the Hong Kong NPC Study Group 0502 trial. A majority (n=559) of patients had undetectable post-treatment plasma EBV DNA on qPCR, while a small proportion had detectable EBV DNA and were either excluded from the trial (n=110) or randomized to the observation arm (n=49) or intervention arm (n=51).

Targeted sequencing was performed on plasma samples (2 mL) collected at 6–8 weeks after treatment with enrichment of EBV DNA molecules using the hybridization-based capture system. For each sample, plasma autosomal and EBV DNA fragments were quantified. This was followed by measurement of the proportions of plasma EBV DNA fragments, as well as size profiling of plasma autosomal and EBV DNA fragments.

Results showed significantly better performance of targeted sequencing vs qPCR in predicting disease progression within 1 year.

Using a low cut-off of plasma EBV DNA proportion of 0.01 percent, sequencing-based analysis demonstrated sensitivity of 88.5 percent and 97.1 percent for predicting local and distant recurrences, respectively, compared with 42.3 percent and 85.3 percent, respectively, for qPCR with plasma EBV DNA >0 copy/mL as cut-off (p=0.0018 for local recurrence, p=0.0078 for distant recurrence). Specificity was 47 percent for sequencing-based analysis and 79.1 percent for qPCR at the respective cut-offs.

At 0.01 percent cut-off for sequencing-based analysis, negative predictive values (NPVs) were 99.1 percent for local recurrence and 99.4 percent for distant recurrence. At >0 copy/mL cut-off for qPCR, the corresponding NPVs were 97.3 percent and 98.2 percent, respectively.

Among samples with undetectable EBV DNA on qPCR (n=559), a wide range of EBV DNA proportions (0–6.4 percent) was found in sequencing-based analysis. When these samples were divided into two subgroups of EBV DNA proportions <0.01 percent (n=182) and ≥0.01 percent (n=377), those with EBV DNA proportions <0.01 percent were found to have significantly higher 5-year rates of overall survival (OS; 93.5 percent vs 85 percent; p=0.001), progression-free survival (PFS; 87.2 percent vs 72.9 percent; p=0.0003), locoregional failure–free survival (LFFS; 90.1 percent vs 76.1 percent; p=0.0002) and distant metastasis–free survival (DMFS; 90.6 percent vs 80.3 percent; p=0.001) vs the other subgroup.

“More importantly, the subgroup of patients with EBV DNA proportions <0.01 percent had significantly better survival outcomes than the whole group of patients with undetectable plasma EBV DNA by qPCR alone [OS, p=0.016; PFS, p=0.007; LRFS, p=0.006; DMFS, p=0.017],” the investigators reported.

When divided into five subgroups of EBV DNA proportions <0.01 percent, ≥0.01–0.1 percent, ≥0.1–1 percent, ≥1–10 percent and ≥10 percent, each 10-fold increase in EBV DNA proportions was associated with hazard ratios (HRs) for OS of 2.51, PFS of 2.43, LRFS of 2.31 and DMFS of 2.53.

“On multivariate analysis, sequencing-based quantitative level of plasma EBV DNA was the independent prognostic factor with the highest HRs for OS, PFS, LRFS and DMFS,” the investigators highlighted.

“These results would facilitate adoption of plasma EBV DNA as a biomarker to individualize adjuvant treatment. We envision that incorporation of sequencing analysis of plasma EBV DNA into routine clinical practice would improve post-treatment management of NPC,” they concluded.