Highlights
In a large prospective multicenter outpatient cohort of 3777 evaluable individuals with clinically suspected nasopharyngeal carcinoma, anti-EBV BNLF2b total antibody (P85-Ab) showed strong overall diagnostic performance, with 93.0% sensitivity and 97.3% specificity.
P85-Ab outperformed traditional Epstein-Barr virus serologic markers, including VCA-IgA, EA-IgA, and EBNA1-IgA, particularly in specificity, where the comparator assays were all below 90%.
P85-Ab retained high sensitivity in asymptomatic individuals and in patients with nonspecific symptoms, suggesting potential value in earlier-stage or diagnostically uncertain outpatient presentations.
A triplet strategy combining P85-Ab, VCA-IgA, and EBNA1-IgA did not improve diagnosis in asymptomatic or nonspecific presentations, but it did increase sensitivity among patients with NPC-specific symptoms.
Background
Nasopharyngeal carcinoma (NPC) is a geographically patterned epithelial malignancy with particularly high incidence in southern China and other endemic regions. Its clinical importance lies not only in cancer mortality, but also in the diagnostic challenge created by its often subtle early manifestations. Patients may present with cervical lymphadenopathy, epistaxis, unilateral serous otitis media, headache, cranial neuropathy, or other local symptoms, but many initially have vague or nonspecific complaints. In outpatient settings, this creates a familiar dilemma: which patients warrant immediate endoscopic and imaging workup, and which biomarkers can most reliably guide triage?
Epstein-Barr virus (EBV) is central to the biology of nonkeratinizing NPC, and EBV-related blood tests have long been used to support diagnosis and screening in endemic areas. However, the performance of conventional serologic markers such as viral capsid antigen immunoglobulin A (VCA-IgA), early antigen immunoglobulin A (EA-IgA), and EBV nuclear antigen 1 immunoglobulin A (EBNA1-IgA) has been variable across settings. Many assays have acceptable sensitivity but suboptimal specificity, which can be problematic in outpatient populations with overlapping benign inflammatory or infectious conditions. False positives can trigger unnecessary endoscopy, imaging, anxiety, and follow-up, while false negatives can delay diagnosis.
The anti-EBV BNLF2b total antibody assay, referred to here as P85-Ab, is a newer biomarker designed to capture the host immune response to an EBV lytic-cycle-associated antigen. The clinical question addressed by this study is highly practical: in real-world suspected NPC populations, does P85-Ab improve diagnostic accuracy compared with existing EBV serologies, and is there any added value in combining it with established antibodies?
Study Design
This investigation was a prospective, multicenter cohort study conducted between April 2021 and March 2024 in outpatient clinics across 5 medical centers in China. The design is notable for its pragmatic clinical orientation: participants were consecutively recruited based on suspicion of NPC in outpatient practice, rather than selected from highly controlled case-control samples that often exaggerate diagnostic performance.
The study enrolled 3795 eligible participants. After excluding individuals with low-quality samples or loss to follow-up, 3777 participants were included in the final analysis. Of these, 1680 had NPC and 2097 did not. The median age was 49.0 years (interquartile range, 37.0-58.0 years), and 65% were male. Median follow-up was 27.2 months, allowing more reliable disease ascertainment among those initially categorized as not having NPC.
P85-Ab was measured using chemiluminescent immunoassay, while VCA-IgA, EA-IgA, and EBNA1-IgA were measured using enzyme-linked immunosorbent assays. The primary endpoints were the sensitivity and specificity of P85-Ab. The study also performed head-to-head comparisons with the other EBV biomarkers and evaluated performance in clinically relevant subgroups, including asymptomatic individuals, those with nonspecific symptoms, and those with NPC-specific symptoms.
Key Results
Overall diagnostic performance
P85-Ab demonstrated the strongest overall balance of sensitivity and specificity among the tested biomarkers. Its sensitivity was 93.0% (95% CI, 91.6%-94.0%), and its specificity was 97.3% (95% CI, 96.5%-97.9%). These figures are clinically impressive in a suspected-disease outpatient cohort, where diagnostic noise is usually substantial.
The comparator biomarkers showed lower performance. Their sensitivities ranged from 60.4% to 93.0%, and all had specificities below 90%. Although the abstract does not provide each assay’s full point estimates in this summary, the stated performance gap is clinically meaningful. In practice, a marker with specificity near 97% is much better positioned to reduce false-positive referrals than one with specificity in the 80% range, especially when used in high-volume outpatient triage.
Performance in asymptomatic and nonspecific presentations
One of the most important findings is that P85-Ab remained highly sensitive outside classic symptomatic disease. Among asymptomatic individuals, sensitivity was 92.0% (95% CI, 85.9%-95.6%). Among those presenting with NPC-nonspecific symptoms, sensitivity was 88.4% (95% CI, 75.5%-94.9%).
These subgroup results matter because early NPC may be silent or clinically ambiguous. Diagnostic tools often perform best in advanced, obvious disease, but what clinicians need most is a biomarker that remains informative when the pretest clinical signal is weak. The preservation of high sensitivity in these groups suggests that P85-Ab may be particularly useful in the front end of the diagnostic pathway, when symptoms alone are insufficiently discriminating.
Combination testing versus P85-Ab alone
The study also assessed whether a triplet-antibody strategy combining P85-Ab, VCA-IgA, and EBNA1-IgA added value. The answer depended on the clinical phenotype.
For asymptomatic individuals and those with nonspecific symptoms, the triplet strategy did not provide additional diagnostic benefit beyond P85-Ab alone. This is an important efficiency finding. If a single assay can deliver high accuracy without materially sacrificing sensitivity, then adding more serologies may increase complexity and cost without improving decision-making.
However, among participants with NPC-specific symptoms, the triplet-antibody strategy improved sensitivity compared with P85-Ab alone: 95.9% (95% CI, 94.8%-96.8%) versus 93.1% (95% CI, 91.7%-94.3%), with P < .001. This is a statistically significant increase. Clinically, the relevance depends on what tradeoff in specificity, cost, and workflow accompanies that gain, details that would be important to review in the full text. Even so, the data support a symptom-stratified approach rather than a one-size-fits-all testing algorithm.
Clinical Interpretation
This study addresses a long-standing problem in EBV-based NPC diagnostics: conventional serologies can be useful, but they have often lacked the specificity needed for confident outpatient decision support. P85-Ab appears to shift that balance favorably. A sensitivity of 93.0% limits missed cases, while a specificity of 97.3% should markedly reduce unnecessary workup among those without cancer.
The strongest translational message is that P85-Ab may function well as a first-line serologic rule-in and rule-out tool in suspected NPC, especially in endemic settings where outpatient clinics evaluate large numbers of patients with low-to-intermediate pretest probability. In asymptomatic individuals or those with nonspecific complaints, the study suggests that P85-Ab alone may be sufficient. In those with more clearly suggestive symptoms, combining P85-Ab with VCA-IgA and EBNA1-IgA could modestly increase sensitivity and potentially reduce the chance of missed disease.
This phenotype-based strategy aligns with modern diagnostic stewardship. Biomarker panels should not automatically be expanded unless they improve outcomes that matter clinically, such as missed cancer, time to diagnosis, number of invasive procedures, or cost per diagnosis. The study’s data imply that broader panels should be reserved for symptom-defined higher-risk subgroups rather than applied universally.
Why BNLF2b May Be Biologically Plausible
Although the abstract is focused on diagnostic performance rather than mechanism, the target antigen is biologically interesting. EBV-associated NPC is characterized by a distinct host-virus interaction, including expression of latent and selected lytic components. BNLF2b is linked to the EBV lytic program, and antibodies against this target may reflect a disease-associated immune signature different from that captured by more traditional antigens such as VCA, EA, or EBNA1. If that signal is more specific to NPC-related viral activity or host response, the superior specificity of P85-Ab becomes biologically plausible rather than merely empirical.
That said, serologic performance can vary by assay platform, antigen design, and population prevalence. The chemiluminescent immunoassay format used for P85-Ab may also contribute to reproducibility and operational scalability, advantages that matter when translating a biomarker from research settings into broad clinical use.
Strengths of the Study
Several design features strengthen confidence in the findings. First, the study was prospective and multicenter, reducing the selection bias common in retrospective or single-center analyses. Second, participants were consecutively recruited from outpatient clinics, making the cohort clinically realistic. Third, the head-to-head comparison against commonly used EBV biomarkers provides direct evidence relevant to practice rather than relying on historical controls. Fourth, the follow-up duration of 27.2 months helps ensure that participants classified as non-NPC were less likely to harbor occult missed disease at baseline.
Another major strength is the attention to symptom-defined subgroups. Diagnostic tests rarely perform identically across all clinical presentations, and the study’s differentiated results are more useful to clinicians than a single pooled accuracy estimate.
Limitations and Cautions
Despite its strengths, the study has limitations that should temper overgeneralization. The cohort was drawn from 5 medical centers in China, an epidemiologically appropriate setting for NPC research but one that may differ from low-incidence regions in referral patterns, background EBV exposure, and pretest probability. Performance characteristics, particularly positive predictive value and negative predictive value, will shift when applied to populations with different NPC prevalence.
The abstract does not detail stage-specific performance, which would be highly relevant because the clinical value of a biomarker is greatest when it improves detection of early-stage disease. Likewise, while improved sensitivity with the triplet strategy in symptom-specific patients is statistically significant, the full clinical tradeoff requires knowledge of any corresponding specificity loss and the consequences for downstream procedures.
Another practical issue is implementation. Laboratories considering adoption will need information on assay standardization, calibration, reproducibility across platforms, quality control requirements, turnaround time, and cost-effectiveness. Those data are not provided in the abstract. Finally, although median follow-up was substantial, the study does not replace histopathology, endoscopy, or imaging. Rather, it informs how serology may best guide who proceeds to those confirmatory evaluations.
Implications for Practice
For clinicians in endemic or high-volume referral settings, the study supports considering P85-Ab as a preferred serologic biomarker for suspected NPC in outpatient practice. A reasonable interpretation of the presented data is as follows:
In asymptomatic individuals with concern for NPC, P85-Ab alone may offer a high-yield, efficient initial test.
In patients with nonspecific symptoms, P85-Ab alone also appears appropriate, with little evidence that routine multi-antibody panels improve diagnostic value.
In patients with NPC-specific symptoms, such as suspicious cervical nodes, unilateral ear symptoms, recurrent epistaxis, or cranial nerve findings, combining P85-Ab with VCA-IgA and EBNA1-IgA may be justified when the clinical goal is to maximize sensitivity.
Even with these data, test selection should remain integrated with endoscopic examination, imaging, and tissue diagnosis. Biomarkers are most valuable when embedded in a structured pathway rather than used in isolation.
Conclusion
This prospective multicenter cohort study provides strong evidence that anti-EBV BNLF2b total antibody is a robust diagnostic biomarker for suspected nasopharyngeal carcinoma in outpatient settings. With 93.0% sensitivity and 97.3% specificity, P85-Ab outperformed conventional EBV serologies and showed particularly attractive specificity, a key advantage for reducing false-positive workup.
The study also suggests a practical symptom-stratified testing model. P85-Ab alone may be sufficient for asymptomatic individuals and those with nonspecific symptoms, whereas a triplet strategy adding VCA-IgA and EBNA1-IgA may be useful in patients with NPC-specific symptoms when sensitivity is prioritized. Future work should confirm generalizability outside endemic regions, define stage-specific performance, and evaluate cost-effectiveness and laboratory implementation. Even so, this study meaningfully advances the evidence base for EBV-guided diagnosis of NPC in real-world practice.
Funding and ClinicalTrials.gov
The abstract provided does not report funding details or a ClinicalTrials.gov registration number. These should be checked in the full JAMA Oncology article before use in formal policy, guideline, or institutional implementation documents.
Citation
Li SC, Li FG, Wu SJ, Tang MZ, Xiao ZZ, Li TD, Xia NS, Mai HQ, Ji MF, Tang LQ, P85-Ab Collaborative Group, Yuan L, Yu X, Li HJ, Lin YY, Liu WL, Wu BH, Xing S, Chen YS, Wu YX, Liu LT, Guo SS, He XF, Li X, Fu MY, Long GX, Tang H, Ma J, Gao TS, Ye JB, Li LJ, Guo QJ, Lv XF, Xiao XH, Lu SL, Zhong J, Li ZM, Song M, Chen SW, Lv X, Xia WX, Hua YJ, Wang L, Yang Q, Zou X, Liang H, Miao JJ, Xie RL, Liu SL, Li XY, Sun XS, Ge SX, Zhang J. Diagnostic Performance of Anti-Epstein-Barr Virus BNLF2b in Suspected Nasopharyngeal Carcinoma. JAMA Oncology. 2026 May 1;12(5):478-487. PMID: 41854586. https://pubmed.ncbi.nlm.nih.gov/41854586/