Abstract
BACKGROUND: HIV-associated lymphomas (HALs) exhibit aggressive features and poorer prognosis compared to HIV-negative lymphomas. However, their molecular and clinicopathological characteristics remain unclear in the antiretroviral therapy (ART) era. METHODS: We retrospectively analyzed 208 lymphoma patients (57 HALs, 151 HIV-negative lymphomas) diagnosed between July 2019 and March 2024. Quantitative immunohistochemistry evaluated expression levels of Ki67, CD10, BCL6, MUM1, BCL2, and MYC. Independent prognostic factors were identified using multivariate Cox regression analysis, and a survival prediction model was validated by receiver operating characteristic (ROC) curve analysis. RESULTS: HALs exhibited significantly higher proliferative activity (Ki67 AOD: 0.92 vs. 0.82, P < 0.001), more advanced disease stages (Ann Arbor stage III/IV: 77.2% vs. 60.0%, P = 0.022), and increased Epstein-Barr virus (EBV) positivity (51.1% vs. 17.9%, P < 0.001). Immunophenotyping revealed a GCB-like phenotype in HALs, characterized by elevated CD10 and BCL6 expression and decreased MUM1 and BCL2 expression. Patients with HALs had significantly shorter survival (median: 32.1 vs. 46.1 months, P < 0.001). Multivariate analysis identified Ki67 AOD (hazard ratio [HR] = 3.04, 95% confidence interval [CI]: 3.85-10.85), International Prognostic Index (IPI) (HR = 9.35, 95% CI: 4.20-20.82), and ART duration (protective, HR = 0.29/year, 95% CI: 0.19-0.45) as independent prognostic factors. The survival model demonstrated strong predictive accuracy (1-year area under the curve [AUC] = 0.831). CONCLUSIONS: HALs exhibit distinct molecular profiles-including elevated EBV infection, a GCB-like phenotype, increased Ki67 AOD, and decreased BCL2 expression-that contribute to significantly poorer survival compared to HIV-negative lymphomas. Integrating Ki67 AOD and IPI scores into prognostic models may enhance individualized prognosis and optimize treatment strategies for HAL patients.