Saliva biopsy of nasopharyngeal carcinoma (NPC) has been developed in our latest study, indicating the application of oral sampling in NPC detection. Further exploration of the potential for self-sampling from the oral cavity is necessary. A total of 907 various samples from oral cavity, including saliva (n = 262), oropharyngeal swabs (n = 250), oral swabs (n = 210), and mouthwash (n = 185), were collected. Epstein‒Barr virus (EBV) DNA methylation at the 12,420 bp CpG site in EBV genome from the repeat-copy W promoter (Wp) region and at the 11,029 bp CpG site in the single-copy C promoter (Cp) region were simultaneously detected in these samples. A significant increase in EBV methylation, no matter at Wp or Cp region, was found in all types of samples from NPC patients. However, EBV DNA methylation in saliva and oropharyngeal swab showed a better diagnostic performance in detecting NPC. The combination of these two sample types and two markers could help to improve the detection of NPC. Our study further explored the optimal self-sampling methods and detection target in the detection of NPC and may facilitate the application of EBV DNA methylation detection in a home-based large-scale screening of NPC.

Background As an oncovirus, EBV is associated with multiple cancers, including solid tumors and hematological malignancies. EBV methylation plays an important role in regulating tumor occurrence. However, the EBV methylation profiles in EBV-associated tumor tissues are poorly understood. Results In this study, EBV methylation capture sequencing was conducted in several different tumor tissue samples, including NPC, EBVaGC, lung LELC and parotid LELC. Besides, EBV capture sequencing and following qMSP were performed on nasopharyngeal brushing samples from NPC and nasal NKTCL patients. Our results showed that the EBV genome among different types of tumors displayed specific methylation patterns. Among the four types of tumors from epithelial origin (NPC, EBVaGC, lung LELC and parotid LELC), the most significant differences were found between EBVaGC and the others. For example, in EBVaGC, all CpG sites within 1,44,189–1,45,136 bp of the EBV genome sequence on gene RPMS1 were hyper-methylated compared to the others. Differently, significant differences of EBV CpG sites, particularly those located on gene BILF2, were observed between NPC and nasal NKTCL patients in nasopharyngeal brushing samples. Further, the methylated level of BILF2 was further detected using qMSP, and a diagnostic model distinguishing NPC and nasal NKTCL was established. The AUC of the model was 0.9801 (95% CI 0.9524–1.0000), with the sensitivity and specificity of 98.81% (95% CI 93.63–99.94%) and 76.92% (95% CI 49.74–91.82%), respectively. Conclusions Our study reveals more clues for further understanding the pathogenesis of EBV, and provides a possibility for distinguishing EBV-related tumor by detecting specific EBV CpG sites.

Background: The nasopharyngeal brush sampling can effectively collect samples from the nasopharynx. The blind brush sampling does not require the guidance of endoscopy, which is favorable for implementation and dissemination in the community. This study explored methylation markers for nasopharyngeal carcinoma (NPC) at both Epstein–Barr virus (EBV) and its host genome levels, aiming to construct a blind brushing diagnostic method. Methods: EBV DNA capture and methylation sequencing and GEO Illumina 450K methylation array data were used respectively for the discovery of EBV and host methylation markers. The diagnostic method was built in training cohort ( n = 347) and validated in an independent validation cohort ( n = 155). Results: A total of 1 EBV methylation marker (BILF2) and 6 host methylation markers (ITGA4, IMPA2, ITPKB, PI9, AMIGO2, and VAV3) were identified. Both EBV and host methylation markers were almost exclusively detected in NPC samples, with negligible detection in control samples. In validation cohort, the diagnostic method that included only the EBV BILF2 marker showed a sensitivity and specificity of 80.22% and 98.44%, respectively. When combining the EBV-derived marker BILF2 with the host-derived marker IMPA2, the diagnostic method’s sensitivity increased to 84.62%, while the specificity remained unchanged (IDI = 4.4%, P = 0.0419). Conclusion: Overall, the blind nasopharyngeal brushing diagnostic method, combining EBV and host methylation markers, showed great potential in NPC detection and could promote its application in nonclinical screening of NPC.

Saliva sampling is a non-invasive method, and could be performed by donors themselves. However, there are few studies reporting biomarkers in saliva in the diagnosis of NPC. A total of 987 salivary samples were used in this study. First, EBV DNA methylation was profiled by capture sequencing in the discovery cohort (n = 36). Second, a q-PCR based method was developed and five representative EBV DNA CpG sites (11 029 bp, 45 849 bp, 57 945 bp, 66 226 bp and 128 102 bp) were selected and quantified to obtain the methylated density in the validation cohort1 (n = 801). Third, a validation cohort2 (n = 108) was used to further verify the differences of EBV methylation in saliva. A significant increase of EBV methylation was found in NPC patients compared with controls. The methylated score of EBV genome obtained by capture sequencing could distinguish NPC from controls (sensitivity 90%, specificity 100%). Further, the methylated density of EBV DNA CpG sites revealed by q-PCR showed a good diagnostic performance. The sensitivity and specificity of detecting a single CpG site (11 029 bp) could reach 75.4% and 99.7% in the validation cohort1, and 78.2% and 100% in the validation cohort2. Besides, the methylated density of the CpG site was found to decrease below the COV in NPC patients after therapy, and increase above the COV after recurrence. Our study provides an appealing alternative for the non-invasive detection of NPC without clinical setting. It paves the way for conducting a home-based large-scale screening in the future.

Detecting EBV DNA load in nasopharyngeal (NP) brushing samples for the diagnosis of nasopharyngeal carcinoma (NPC) has attracted widespread attentions. Currently, NP brush sampling mostly relies on endoscopic guidance, and there are few reports on diagnostic markers suitable for nonguided conditions (blind brush sampling), which is of great significance for extending its application. One hundred seventy nasopharyngeal brushing samples were taken from 98 NPC patients and 72 non-NPC controls under the guidance of endoscope, and 305 blind brushing samples were taken without endoscopic guidance from 164 NPC patients and 141 non-NPC controls (divided into discovery and validation sets). Among these, 38 cases of NPC underwent both endoscopy-guided NP brushing and blind brushing. EBV DNA load targeting BamHI-W region and EBV DNA methylation targeting 11029 bp CpG site located at Cp-promoter region were detected by quantitative polymerase chain reaction (q-PCR). EBV DNA load showed good classification accuracy for NPC in endoscopy-guided brushing samples (AUC = 0.984). However, in blind bushing samples, the diagnostic performance was greatly reduced (AUC = 0.865). Unlike EBV DNA load, the accuracy of EBV DNA methylation was less affected by brush sampling methods, whether in endoscopy-guided brushing (AUC = 0.923) or blind brushing (AUC = 0.928 in discovery set and AUC = 0.902 in validation set). Importantly, EBV DNA methylation achieved a better diagnostic accuracy than EBV DNA load in blind brushing samples. Overall, detection of EBV DNA methylation with blind brush sampling shows great potential in the diagnosis of NPC and may facilitate its use in nonclinical screening of NPC.