Determining the risk of developing EBV­-associated lymphoma in HIV-­infected patients

Abstract

Objective — to increase the efficiency of predicting the development of EBV-associated lymphomas in patients with HIV infection.
 Materials and methods. In order to improve the prediction of the development of Epstein—Barr associated lymphomas in HIV-infected patients, a comparative analysis of various indicators (clinical-epidemiological, laboratory, serological, etc.) was carried out in 57 HIV patients who had clinical and laboratory confirmation of the presence of co-infection with the Epstein—Barr.Of the 57 patients, manifestations of primary CNS lymphoma were registered in 7 patients (12.3 %), another 1 patient was diagnosed with B-cell large cell lymphoma of the frontal sinus, centroblastic variant (1.8 %), in 1 patient — Burkitt’s lymphoma with lesions of the cervical lymph nodes (1.8 %). To clarify the main trends in the development of lymphomas in patients with HIV and EBV coinfection, a detailed analysis was carried out in two groups: the main group consisted of 9 patients with lymphomas, the remaining 48 patients without detected neoplasms formed the comparison group.
 Results and discussion. To improve the efficiency of the forecast, multifactorial logistic regressions were constructed, taking into account not only the independent, but also the joint influence of the considered risk factors. To do this, the sum of the scores for each observation was calculated using the corresponding predictive coefficient rank qualification created using the Wald analysis. Based on a multivariate prognostic model, an algorithm was created to determine the risk of developing EBV-associated lymphomas in patients co-infected with EBV and HIV.
 Conclusions.The created algorithm for determining the risk of developing EBV-associated lymphomas in patients with HIV makes it possible to identify patients with different risks of developing lymphomas under conditions of infection with EBV and HIV, which further creates the possibility for predicting an unfavorable course of HIV infection.The proposed algorithm has a high predictive efficiency, which makes it possible to determine the risk at the individual level and lays the foundation for optimizing the diagnosis of Epstein—Barr virus infection in HIV-infected patients.