Polygenic Risk Scores for Prediction of Subclinical Coronary Artery Disease in Persons With Human Immunodeficiency Virus (HIV): The Swiss HIV Cohort Study
Isabella C Schoepf, Christian W. Thorball, Helen Kovari, Bruno Ledergerber, Ronny R. Buechel, Alexandra Calmy, Rainer Weber, Philipp A. Kaufmann, René Nkoulou, Johannes M. Schwenke, Dominique L. Braun, Jacques Fellay, Philip Tarr, for the Swiss HIV Cohort Study, Irène A. Abela, Karoline Aebi‐Popp, A Anagnostopoulos, M Battegay, Enos Bernasconi, Dominique L. Braun, H C Bucher, Alexandra Calmy, Matthias Cavassini, Angela Ciuffi, G Dollenmaier, Matthias Egger, Luigia Elzi, J Fehr, Jacques Fellay, H Furrer, C A Fux, Huldrych F. Günthard, Anna Hachfeld, David Haerry, Barbara Hasse, Hans H. Hirsch, Marc Hoffmann, Irène Hösli, M. Huber, Christian R. Kahlert, Laurent Kaiser, Olivia Keiser, Thomas Klimkait, Roger D. Kouyos, Helen Kovari, Katharina Kusejko, G Martinetti, Begoña Martínez de Tejada, Catia Marzolini, Karin J. Metzner, N Müller, Johannes Nemeth, Dunja Nicca, P Paioni, Giuseppe Pantaleo, Matthieu Perreau, Alan Rauch, Patrick Schmid, R Speck, M Stöckle, Philip Tarr, Alexandra Trkola, Gilles Wandeler, Sabine Yerly
Abstract
BACKGROUND: In people with human immunodeficiency virus (HIV) (PWH), individual polygenic risk scores (PRSs) are associated with coronary artery disease (CAD) events. Whether PRSs are associated with subclinical CAD is unknown. METHODS: In Swiss HIV Cohort Study participants of European descent, we defined subclinical CAD as presence of soft, mixed, or high-risk plaque (SMHRP) on coronary computed tomography (CT) angiography, or as participants in the top tertile of the study population's coronary artery calcium (CAC) score, using noncontrast CT. We obtained univariable and multivariable odds ratios (ORs) for subclinical CAD endpoints based on nongenetic risk factors, and validated genome-wide PRSs built from single nucleotide polymorphisms associated with CAD, carotid intima-media thickness (IMT), or longevity in the general population. RESULTS: We included 345 genotyped participants (median age, 53 years; 89% male; 96% suppressed HIV RNA); 172 and 127 participants had SMHRP and CAC, respectively. CAD-associated PRS and IMT-associated PRS were associated with SMHRP and CAC (all P < .01), but longevity PRS was not. Participants with unfavorable CAD-PRS (top quintile) had an adjusted SMHRP OR = 2.58 (95% confidence interval [CI], 1.18-5.67), and a CAC OR = 3.95 (95% CI, 1.45-10.77) vs. bottom quintile. Unfavorable nongenetic risk (top vs. bottom quintile) was associated with adjusted SMHRP OR = 24.01 (95% CI, 9.75-59.11), and a CAC-OR = 65.07 (95% CI, 18.48-229.15). Area under the receiver operating characteristic curve increased when we added CAD-PRS to nongenetic risk factors (SMHRP: 0.75 and 0.78, respectively; CAC: 0.80 and 0.83, respectively). CONCLUSIONS: In Swiss PWH, subclinical CAD is independently associated with an individual CAD-associated PRS. Combining nongenetic and genetic cardiovascular risk factors provided the most powerful subclinical CAD prediction.