CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Poster Abstracts

866 WITHDRAWN / INTENTIONALLY UNASSIGNED 867 CLUSTER SURVEILLANCE OF FRENCH PRIMARY INFECTIONS: TOWARD A MORE VIRULENT CRF02_AG? Benoit Visseaux 1 , Lambert Assoumou 1 , Mary Anne Trabaud 2 , Brigitte Montes 3 , Laurence Bocket 4 , Samira Fafi-Kremer 5 , Marc Wirden 1 , Corinne Amiel 6 , Anne De Monte 7 , Karl Stefic 8 , Camille Tumiotto 9 , Anne Maillard 10 , Diane Descamps 1 , Marie-Laure Chaix Baudier 1 , for the AC43 ANRS Resistance Group 1 INSERM, Paris, France, 2 CHU de Lyon, Lyon, France, 3 CHU de Montpellier, Montpellier, France, 4 CHU de Lille, Lille, France, 5 INSERM, Strasbourg, France, 6 Tenon Hospital, Paris, France, 7 CHU de Nice, Nice, France, 8 INSERM, Tours, France, 9 CHU de Bordeaux, Bordeaux, France, 10 CHU de Rennes, Rennes, France Background: Molecular epidemiology can be used to identify large recent transmission clusters (RTC) and describe core transmitters that fuel a large proportion of transmissions. We analyzed such RTC among primary infected patients (PHI) diagnosed in France in 2014-2016. Methods: Protease and reverse transcriptase sequences were obtained from 1121 patients included between 2014 and 2016 from 46 centers. Phylogenetic trees were built by approximate maximum likelihood using FastTree to identify RTC (max genetic distance <4.5%, branch support value >95%). Results: Most patients were men (90%), MSM (70%), born in France (70%) or Sub-Saharan Africa (6.6%), infected mostly by B (56%) or CRF02_AG (20%) clades. CRF02_AG tended to be increasingly represented across years (from 17 to 22%) and large (>3 patients) RTC (Table). Compared to patients infected by subtype B, patients infected by CRF02_AG presented a lower proportion of MSM (59 vs 78%, p<0.001), of individual born in France (67 vs 75%, p=0.02), higher viral loads (VL) (median at 5.83 log10 copies/mL [IQR: 4.96-6.60] vs 5.40 [4.66-6.26], p=0.004) and lower CD4 cell counts (463 cells/mm3 [25-903] vs 514 [1-1028], p=0.004). When analyzing patients born in France separately, CRF02_AG still presented higher VL than subtype B (5.79 vs 5.42 log10 copies/ mL, p=0.012). Overall, 457 (41%) patients were included in RTC including 214 (47%) in 106 small (<4 patients) and 243 (53%) patients in 39 large RTC (from 4 to 14 patients). Paris area appeared as a hub for transmission with 31/39 large RTC including ≥1 patient from this area. RTC-patients were younger and more frequently MSM than non-RTC-patients (p<0.001). Most large RTC sustained active transmissions over the whole study period. Four large clusters were identified with transmitted drug resistance mutation(s) (T215S, L74M, K103N and L76V+L90M) but none achieved sustainable transmission of these mutations throughout their cluster. Conclusion: This study highlights the important role of RTC achieving transmission throughout France with a large hub in Paris area. CRF02_AG is actively spreading among large RTC, participating to the epidemiological shift from B to CRF02_AG in France. CRF02_AG is also associated to higher VL among patients born in France, suggesting a higher virulence than subtype B. The increasing number of large RTC identified highlights the need for nationwide

865 HIV DYNAMICS IN THE MOST AFFECTED AREA OF EUROPE: A TALE OF 2 COUNTRIES Lise Marty 1 , Liis Lemsalu 2 , Dominique Costagliola 1 , Kaire Vals 2 , Ruta Kaupe 3 , Indra Linina 3 , Inga Upmace 3 , Kristi Rüütel 2 , Anda Ķīvīte 3 , Virginie Supervie 1 , for the HERMETIC study group 1 INSERM, Paris, France, 2 National Institute for Health Development, Tallinn, Estonia, 3 Riga Stradiņš University, Riga, Latvia Background: In 2016, Latvia and Estonia continued to have the highest rate of new HIV diagnoses in Europe (1.85 and 1.74 per 10000, respectively). Both countries experienced an HIV outbreak among people who inject drugs (PWID) in the early 2000s. Methods: Data from 2000-2016 for persons newly diagnosed with HIV in Latvia and persons newly appearing with HIV in health care registries in Estonia were used in a clinical-stage based back-calculation model to estimate: HIV incidence, time from infection to diagnosis and undiagnosed HIV prevalence. Population size estimates were calculated using national statistics and studies on sexual behavior and drug use. Statistical comparisons were carried out using Mann-Whitney test for incidence and undiagnosed prevalence rates, and using two-sided Kolmogorov-Smirnov test for the distribution of times between infection and diagnosis. Results: In 2016, HIV incidence was twice as high in Latvia than in Estonia (3.5/10000 vs 1.9/10000, p<0.05). Between 2010-2016, HIV incidence decreased in Estonia but increased in Latvia (average annual change of -9.0% and +6.2%, respectively; Table). The incidence decreased for all exposure groups in Estonia and increased for most in Latvia, especially for women and men who have sex with men (MSM). Between 2012-2016, time to diagnosis took longer in Latvia than in Estonia (3.9 vs 3.4 years, p<0.05). In Latvia, getting diagnosed tended to take longer for heterosexual men and MSM than for PWID and heterosexual women (respectively 4.8 and 4.4 vs 3.4 and 3.7 years). A similar trend was observed in Estonia. Undiagnosed prevalence rate was higher in Latvia than in Estonia (13.7/10000 vs 10.3/10000, p<0.05). In 2016, PWID were the most affected population in terms of rates of new and undiagnosed infections in both countries, but the vast majority of new and undiagnosed number of infections occurred among heterosexuals and MSM. In Latvia, 61% of new infections and 65% of undiagnosed infections were among heterosexuals and MSM. In Estonia, they were 85% and 83%, respectively. Conclusion: For the first time, we show stark differences in the HIV epidemic status in the neighboring countries most affected by HIV in Europe. Our results suggest that Estonia has started turning the tide of its epidemic while in Latvia it remains very active. Finding men and women who have acquired HIV sexually is one of the biggest challenges in ending HIV in these originally injection drug use driven epidemics.

CROI 2019 337