CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

603 LYMPHOMA IN HIV-2–INFECTED PATIENTS IN CART ERA: A 15-PATIENT SINGLE-CENTER SERIES Anne-Marie Ronchetti 1 , Sophie Matheron 2 , Lionel Galicier 1 , Florence Damond 2 , Nadia Mahjoub 1 , Amel Besseghir 3 , Veronique Meignin 1 , Francois Simon 1 , Eric Oksenhendler 1 , Laurence Gerard 1 1 Saint Louis Hosp, Paris, France, 2 Bichat-Claude Bernard Hosp, Paris, 3 INSERM, Bordeaux, France Background: In France HIV-2 infection represents 1.6% of all HIV infections. Despite difference in pathogeniciy and disease progression between HIV-2 and HIV-1 infection, similar patterns of opportunistic complications are reported at similar CD4 level. However, although lymphoma occurs in 4 to 5% of HIV-1 patients (pts) and is a major cause of mortality, this complication has been rarely reported in HIV-2-infected pts. Methods: Patients were recruited from an ongoing prospective single-center cohort of HIV-lymphoma. Characteristic of HIV-2 pts were compared with HIV-1 pts included in the cohort >1996. Prevalence and incidence of HIV-2-lymphoma were evaluated from a national prospective multicentric French cohort of HIV-2 patients, including 1068 pts (ANRS-CO5). Results: Among the 949 pts included in the HIV-lymphoma cohort, 15 were HIV-2-infected (1.6%), close to the 2% prevalence of HIV-2 infection in our institution (60/3000). In the ANRS CO5 cohort, the prevalence of lymphoma was 1.7% and the incidence 0.9/1000 PY. Compared to HIV-1 pts, HIV-2 pts were older and more pts were originated fromWest Africa (Table). At the time of lymphoma diagnosis, median CD4 cell count was similar in both groups, but more HIV-2 pts had an undetectable viral load. Clinical presentation of lymphoma was aggressive in both groups. Histologic subtype was HL in 18% of HIV-2 and 13% of HIV-1 pts. Among NHL, diffuse large B-cell lymphoma and Burkitt lymphoma were the most frequent histologic subtype in both groups. All but one HIV-2 pt received intent-to-treat chemotherapy, adapted to histologic subtype. Concomitant cART was maintained or introduced in all pts. Complete remission was achieved in 60% of HIV-2 and 73% of HIV-1 pts. The median overall survival (OS) was lower in HIV-2 pts (16 mths) compared to HIV-1 pts (12 yrs). The cause of death in HIV-2 pts was lymphoma (6) and treatment toxicity (4), no pt died from AIDS. Conclusion: This is the first series of lymphoma in HIV-2 pts. The incidence was close to the incidence of HIV-1-lymphoma in France (0.8/1000 for NHL and 1.2/1000 PY for HL). HIV-2 pts developed lymphoma at the same CD4 level than HIV-1 pts, but a higher proportion of pts had controled HIV infection. Despite some differences in histological subtypes, clinical presentation of lymphoma was close to HIV-1. However, although all pts received adapted chemotherapy and concomitant cART, HIV-2-lymphoma displays an unexpected poor survival, with median OS of 16 mths.

Poster and Themed Discussion Abstracts

604 EPSTEIN-BARR VIRUS DNA LOAD: A POSSIBLE PREDICTOR OF AIDS-RELATED LYMPHOMA Simone H. Hijlkema , Jeroen J. van Kampen, David van de Vijver, Ineke van der Ende Erasmus Univ Med Cntr, Rotterdam, Netherlands

Background: After the introduction of combined antiretroviral therapy, lymphoma remains the main cause of AIDS-related death in HIV-infected adults. AIDS-related lymphoma is associated with Epstein-Barr virus (EBV). Patients with high EBV viral loads seemmore at risk for developing AIDS-related lymphoma. The aim of this study was to investigate the relationship between the level of serum/plasma EBV DNA load and non-Hodgkin lymphoma. Methods: We included HIV-positive patients with stored EBV DNA serum and plasma samples between January 2004 and October 2015 in an academic hospital in the Netherlands. We compared EBV DNA load between patients with and without non-Hodgkin lymphoma. To improve power of the study we randomly selected four controls for each case. EBV DNA was measured using a quantitative real-time Taqman PCR. The EBV DNA load was 10 log transformed. Results: We identified 157 patients of whom 31 were diagnosed with non-Hodgkin lymphoma. Patients had a median age of 43, were predominantly male (80%) and the majority of the patients was born in Europe (56%). Patients had a median CD4 cell count of 100 cells/µl (interquartile range: 30; 265) and nadir CD4 cell count of 60cells/µl (interquartile range: 20; 176. Median HIV RNA load, after exclusion of patients with undetectable viral loads (n=47), was 100000 copies/ml (interquartile range: 7778; 160500). The geometric mean of the peak EBV DNA load measured before onset of non-Hodgkin lymphoma was 11614 IU/ml in the lymphoma group versus 242 IU/ml in the control group (P<0.001). There were 3 patients in the lymphoma group and 60 patients in the control group who had a peak EBV DNA load below 50 IU/ml. Patients with an EBV DNA load greater or equal to 50 IU/ ml have an increased risk of developing non-Hodgkin lymphoma compared to patients with an EBV DNA load below 50 IU/ml (odds ratio 8.75; 95% confidence interval 2.53; 30.29). Conclusion: This study showed that patients who developed non-Hodgkin lymphoma had higher EBV DNA loads prior to the disease compared to patients who did not develop non-Hodgkin lymphoma. In addition, patients of whom peak EBV viral loads were greater or equal to 50IU/ml had approximately a nine times increased risk of developing non- Hodgkin lymphoma. High blood EBV DNA load can be a potential predictive biomarker to detect non-Hodgkin lymphomas in an earlier stage. 605 THE IMMUNE ACTIVATION PROFILE LINKED TO KAPOSI SARCOMA IN HIV-1–POSITIVE ADULTS Christina Psomas 1 , Mehwish Younas 2 , Jacques Reynes 1 , Pierre Portales 3 , Jean-François Eliaou 3 , Pierre Corbeau 2 1 Univ Hosp of Montpellier, Montpellier, France, 2 Inst of Human Genetics, Montpellier, France, 3 Univ Hosp of Montpellier, Montpellier, France Background: Kaposi sarcoma (KS), a vascular neoplasm, still remains a frequent cancer even in efficiently treated persons living with HIV with a satisfactory degree of immune restoration. Human herpes virus 8 is necessary, but not sufficient to cause KS. In in vitro and in animal models, inflammatory cytokines have been shown to induce KS lesion

CROI 2017 256