Prognostic implications of HPV-related nasopharyngeal carcinoma in non-endemic regions: A meta-analysis.

Abstract

e18147 Background: The cancerogenic role of the Epstein Barr Virus (EBV) is well established in endemic non-keratinizing nasopharyngeal carcinoma (NPC). Nearly 10% of non-endemic (NE) NPCs are keratinizing squamous cell carcinomas, scarcely associated with EBV infection and more strongly linked to tobacco exposure. Some EBV-ve NPCs are HPV+ve, with limited data of its prognostic role. Methods: We performed a meta-analysis (MA) from published studies that included NE EBV-ve/HPV+ve NPC with available data regarding overall survival (OS) and a direct comparison with NPC’s subgroups. Hazard ratios and 95% confidence intervals for OS were extracted from each study and harmonized to a common reference group. Pooled estimates were obtained using random-effects MA, and between-study heterogeneity was quantified using the I² statistic . Results: We retrieved 5 studies including a total of 1084 subjects. In four studies the EBV+ve/HPV-ve NPCs (group A) were compared to EBV-ve/HPV+ve NPCs (group B) and to EBV-ve/HPV-ve NPCs (group C). Group B and C comparison was available in two studies. Group C was associated with significantly worse OS compared to group A in a random-effects meta-analysis (pooled HR = 2.30, 95% CI 1.51–3.52; I²=0%, p<0.001). Group C showed also a significantly worse OS compared to group B (pooled HR = 2.08, 95% CI 1.26–3.44; p=0.004), with no evidence of between-study heterogeneity (I²=0%). Finally, the comparison between group A and B showed a positive trend in favour of EBV+ve NPC, with no statistical significance (pooled HR = 1.41, 95% CI 0.87–2.28, I²=0%, p=0.16), indicating comparable survival outcomes between virus-related NPCs. In table 1 are summarized trials characteristics and the pooled analyses. Conclusions: Our analysis indicates that in NE EBV-ve NPCs a distinct HPV+ve subpopulation display a better OS relative to HPV-ve and a similar OS relative to EBV+ve, supporting HPV testing in all EBV-ve NPCs. Prospective studies are warranted to further validate these findings. Studies included in the Meta-analysis, respective hazard-ratio and p value and pooled analyses. Study Total n° of patients pvalue Outcome Reference Contrast Hazard Ratio (HR) 95% CI Stenmark et al 62 0.390.39 OS AA BC 1.831.26 [0.71, 4.72][0.37, 4.25] Wu et al 78 0.110.0030.414 OS AAB BCC 2.413.751.55 [0.82, 7.11][1.59, 8.84][0.54, 4.45] Ruuskanen et al 150 0.0050.03 OS AB CC 2.272.22 [1.28, 4.01][1.06, 4.76] Huang et al 451 0,73 OS A B 0.85 [0.33, 2.17] Verma et al 343 0.640.61 OS AA BC 1.231.46 [0.51, 2.99][0.33, 4.39] Pooled Analyses Groups Comparison: A vs C Pooled HR: 2.30 95% CI 1.51, 3.52 p<0.001 Groups Comparison: B vs C Pooled HR:2.08 95% CI 1.26, 3.44 p=0.004 Groups Comparison: A vs B Pooled HR: 1.41 95% CI 0.87, 2.28 p=0.16 Legend:A: EBV+ve/HPV-ve. B: EBV-ve/HPV+ve. C: EBV-ve/HPV-ve.