O3-17. Decreasing HIV-1 incidence and undiagnosed HIV-1 cases in Sweden based on multiple biomarker estimate of infection times

Jan Albert [1,2], Federica Giardina [3,4], Ethan Romero-Severson [4], Maria Axelsson [5], Veronica Svedhem [6,7] Thomas Leitner [4], Tom Britton [5]
Affiliates: [1] Department of Microbiology, Tumor and Cell Biology, Karolinska Institute, Stockholm, Sweden. [2] Department of Clinical Microbiology, Karolinska University Hospital, Stockholm, Sweden. [3] Department of Mathematics, Stockholm University, Stockholm, Sweden. [4] Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, Los Alamos, NM. [5] Public Health Agency of Sweden, Solna, Sweden. [6] Unit of Infectious diseases, Department of Medicine Huddinge, Karolinska Institute, Stockholm, Sweden. [7] Department of Infectious Diseases, Karolinska University Hospital, Stockholm, Sweden.

Background
Accurate estimation of time of infection (TI) for persons living with HIV-1 (PLHIV) is important for understanding the epidemiology of HIV and for informing public health decisions. We developed a multiple biomarker method of estimating TI for PLHIV that we use to estimate HIV-1 incidence and proportion of undiagnosed PLHIV in Sweden 2010-2015.

Methods
A Bayesian non-linear mixed effects model based on three biomarkers (CD4 counts, BED serology and polymorphisms in HIV-1 pol sequences) was developed to estimate TI in the three main transmission groups: men who have sex with men (MSM), intravenous drug users (IDUs) and heterosexually infected persons (HET). The model was trained on longitudinal data from 31 treatment-naïve patients with well-defined infection times, and subsequently used to estimate TI for 1,357 patients diagnosed in Sweden during 2003-2010. The model was extended to include all patients in Sweden and the time-distribution between arrival in Sweden and diagnosis for immigrants.

Results
The multiple-biomarker model was more accurate than single biomarkers (mean absolute error 1.01 vs ≥ 1.95). We inferred that domestic infections decreased from 280 (95% CI 256-306) in 2010 to 123 (95% CI 105-139) persons (p<0.01) in 2015. Addressing the UNAIDS 90-90-90 target, we estimate that 813 (95% CI 780-862) PLHIV were undiagnosed in 2015, representing a proportion of 10.8% (95% CI 10.4-11.3%) of all PLHIV.

Conclusions
HIV-1 incidence in Sweden was estimated to have decreased and the proportion of diagnosed PLHIV was estimated to be close to the 90% UNAIDS target.