Investigators analyzed eight primary conjunctival squamous cell carcinomas (cSCC) drawn from two cohorts: five tumors from Nigeria (patients with HIV) and three tumors from Japan (patients without HIV). The team reported that Epstein-Barr virus-encoded RNA detected by EBER-ISH localized to epithelial tumor cells in four of the five Nigerian specimens. In contrast, none of the three Japanese tumors showed EBER signal in tumor epithelium. This pattern, observed across small cohorts from distinct geographic and clinical backgrounds, suggests a possible link between EBV presence and HIV-associated cSCC in the studied samples. The investigators emphasized that their data represent a limited series and were presented as a preprint, not yet peer reviewed.
The authors applied three molecular approaches to interrogate viral presence: EBER in situ hybridization to localize EBV RNA within tissue, EBV DNA PCR to detect viral genomes, and virome analysis of RNA sequencing data to profile viral transcripts. EBV DNA and EBV RNA were each detected in three Nigerian tumors. The report states that every tumor with positive EBER staining was corroborated by at least one additional assay (either DNA PCR or RNA virome evidence). Investigators reported concordant multi-assay support for EBER-positive tumors, but the abstract does not provide the full mapping of which tumors were positive by which specific assays beyond those counts. The source did not list cycle thresholds, sequencing read counts, or assay sensitivity metrics in the abstract.
To test whether EBV can infect and alter conjunctival epithelial cells, the study team infected human telomerase reverse transcriptase-immortalized conjunctival epithelial cells with an Akata strain engineered to express GFP (Akata-GFP-EBV). Infected cultures expressed both latency-associated and lytic viral genes, the authors reported. The in vitro model therefore demonstrated that conjunctival epithelial cells are permissive for EBV expression programs spanning latency and lytic cycles. The investigators compared three infected cultures with three control (uninfected) cultures and identified transcriptional differences between the groups; the abstract reports identification of 96 different—but the remainder of that sentence and the detailed list of altered genes, pathways, or effect sizes were not provided in the abstract. The source did not report specific gene names, fold-changes, pathway enrichment results, or functional assays validating the transcriptional changes in the abstract text.
The preprint provides paired clinical and experimental signals that link Epstein-Barr virus to several HIV-associated cSCC specimens and shows that EBV can induce viral gene expression and alter the transcriptome of immortalized conjunctival epithelial cells. However, the abstract omits several critical details that readers and clinicians will want to see before translating these observations into mechanistic conclusions. The source did not report the full transcriptional signature beyond the fragmentary phrase "identified 96 different...", so the identity of differentially expressed genes, their directions of change, statistical thresholds, and enrichment analyses are not available in the abstract. The authors also did not provide assay-level metrics (for example, PCR thresholds or RNAseq read depth) or clinical metadata for the tumor cases in the abstract; those details may appear in the full preprint but were not in the abstract text.
Additional open questions left by the reported material include whether EBV presence represents a causal contributor to tumorigenesis or a passenger infection in an immune-suppressed niche, how reproducible the viral-tumor association is in larger, geographically diverse cohorts, and which host transcriptional pathways are consistently altered by epithelial EBV infection. The investigators documented multi-assay viral detection and an experimental infection model, yet the abstract does not present longitudinal clinical outcomes, functional validation of candidate genes, or mechanistic experiments linking specific viral factors to oncogenic processes. Those items remain unanswered in the abstract and should be sought in the full manuscript and subsequent peer-reviewed publications.
Limitations explicitly apparent from the abstract include the small case series size, the geographic and HIV-status clustering of the cohorts, and incomplete reporting of the transcriptional profiling results in the abstract. Readers should treat these findings as preliminary and review the full preprint for methods, complete results, and the authors' stated limitations and discussion.