The use of combination antiretroviral therapy (CART) has improved the prognosis of HIV-related diffuse large B cell lymphoma (DLBCL). In the pre-CART era, HIV-infected individuals usually presented DLBCL of primary extranodal origin and advanced stage, unlike non-HIV-infected individuals. In the CART era, studies conducted before the introduction of immunochemotherapy pointed outcomes of HIV-infected patients to be approaching those of non-HIV-infected patients. Recently, studies have reported the lack of influence of HIV-infection in the prognosis of Burkitt’s lymphoma treated with immunochemotherapy. Nevertheless, there is scarce information comparing the clinical presentation and outcome of DLBCL treated with immunochemotherapy between non-HIV-infected patients and HIV-infected patients in CART era.


We retrospectively studied two series of DLBCL patients treated with RCHOP; 81 HIV-infected patients included in a Spanish multicentre trial and 84 non-HIV-infected patients diagnosed in our institution between 2002 and 2010. Demographic, HIV-infection, and DLBCL data on each case were collected. DLBCL cases were classified in nodal and extranodal according to a previously reported definition (JCO 2005). Continuous and categorical variables are presented using descriptive statistics. Survival analyses were performed using the Kaplan-Meier method, and compared using the log-rank test. P-values of less than 0.05 were considered statistically significant.


The median follow-up of HIV-infected patients was 6.5 years and of non-HIV-infected patients was 4.6 years. Non-HIV-infected patients were older at the time of DLBCL diagnosis than HIV-infected patients, median age (range) 62 years (24-80) and 44 years (24-74) respectively; (P<0.001). HIV-infected series had a higher percentage of males (81%) than non-HIV-infected (52%); (P<0.001). There were differences between HIV-infected patients and non-HIV-infected patients regarding the following features: B-symptoms (53% vs. 26%; P<0.001), III and IV Ann Arbor stages (71% vs. 52%; P=0.011), and ECOG performance status higher than 2 (51% vs. 26%; P=0.001). There was a trend for more HIV-infected patients (61%) to have high LDH levels at DLBCL diagnosis than non-HIV-infected patients (45%); P=0.057. On the other hand, the percentage of patients with intermediate-high/high International Prognostic Index, two or more extranodal involvement, and bulky disease was similar in both HIV-infected and non-HIV-infected patients. No differences were observed between HIV-infected and non-HIV-infected series in terms of nodal and extranodal presentation. HIV-infected patients were classified according to the time of CART onset; before immunochemotherapy (69%) and concomitantly with RCHOP (31%). More patients receiving CART prior to immunochemotherapy presented undetectable viral load at lymphoma diagnosis (43%) than patients to whom CART was initiated with DLBCL treatment (4%); P=0.001. Time of CART onset had no effect on the presentation of the remaining clinical features, as well as in CD4-lymphocyte counts at DLBCL diagnosis. Non-HIV-infected patients tended to have a better complete response (CR) rate than HIV-infected patients (81% vs. 69%; P=0.071). Within HIV-infected patients the time of CART onset did not influence the achievement of CR to RCHOP. The overall survival (OS) was better in non-HIV-infected patients (5-year OS 74%; 95%CI: 64-84%) than in HIV-infected patients (5-year OS 56%; 95%CI: 45-67%); P=0.009. No differences in disease free survival (DFS) were observed between HIV-infected (5-year DFS 72%; 95%CI: 60-84%) and non-HIV-infected patients (5-year DFS 73%; 95%CI: 62-84%). In HIV-infected patients time of CART onset neither had impact in OS nor in DFS. Nodal and extranodal DLBCL cases showed similar OS and DFS.


In the CART era, DLBCL of HIV-infected patients still have aggressive features but the incidence of extranodal involvement equalizes that of non-HIV-infected patients. The prognosis of DLBCL treated with RCHOP is worse in HIV-infected than in non-HIV-infected patients.

Supported in part by grants EC11-041 and RD12/0036/0029 RTICC from Instituto Carlos III, Spain.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.