DOI: 10.1161/circulationaha.126.079459 ISSN: 0009-7322
Association of Clonal Hematopoiesis With Silent Brain Lesions and Cognitive Decline in Patients With Atrial Fibrillation
Pascal B. Meyre, Hyo-Jeong Ahn, Carolin A. Ehlert, Tsai-Sang Dederichs, Stefanie Aeschbacher, Andreas S. Müller, Jürg H. Beer, Tobias Reichlin, Nicolas Rodondi, Elisavet Moutzouri, Steffen Schneider, Uwe Zeymer, Constantin von zur Mühlen, Dennis Wolf, Dirk Westermann, Giorgio Moschovitis, Leo H. Bonati, Stefan Osswald, Felix Mahfoud, Christine S. Zuern, Michael Chong, Guillaume Paré, David Conen, Eue-Keun Choi, Michael Kühne, Ingo Hilgendorf
BACKGROUND:
Clonal hematopoiesis of indeterminate potential (CHIP) has been associated with cardiovascular disease, but its relationship with silent brain lesions and cognitive decline in patients with atrial fibrillation (AF) is unclear.
METHODS:
In this prospective, multicenter cohort study, we included 1572 patients with AF enrolled between 2014 and 2017. All participants underwent deep-targeted sequencing for CHIP-associated sequence variations, brain magnetic resonance imaging, and standardized cognitive assessments. CHIP carrier status (overall,
DNMT3A
carrier, and non-
DNMT3A
carrier) was evaluated in relation to prevalent and incident silent brain lesions. Silent brain lesions were defined as large noncortical or cortical infarcts, small noncortical infarcts, microbleeds, and white matter lesions (WMLs), and were assessed at baseline and after 2 years with magnetic resonance imaging. Cross-sectional analyses were externally validated in a cohort of 199 patients with AF. Cognitive function was evaluated through 7 years using the Montreal Cognitive Assessment. Associations were analyzed using multivariable-adjusted logistic and linear regression models.
RESULTS:
Among 1572 patients with AF (mean±SD age, 72.5±8.3 years; 26% women), 342 (22%) carried CHIP sequence variations, most commonly in
DNMT3A
(49.4%) and
TET2
(28.3%). CHIP carriers had significantly higher odds of cerebral microbleeds (odds ratio [OR], 1.45 [95% CI, 1.09 to 1.93]) and WMLs (OR, 1.56 [95% CI, 1.20 to 2.04]) at baseline, particularly with non-
DNMT3A
sequence variations (microbleeds: OR, 1.79 [95% CI, 1.25 to 2.55]; WMLs: OR, 1.80 [95% CI, 1.25 to 2.62]). A similar trend was observed in an independent AF patient cohort from Korea limited to
TET2
sequence variations. Presence of multiple CHIP sequence variations further increased the odds of microbleeds (OR, 1.36 [95% CI, 1.10 to 1.69]) and WMLs (OR, 1.41 [95% CI, 1.15 to 1.75]), and large CHIP clones (variant allele frequency >10%) were associated with higher WML volumes. At 2-year follow-up, CHIP was associated with new cerebral microbleed counts and greater WML volume. Over 7 years, CHIP carriers exhibited greater cognitive decline, with a lower Montreal Cognitive Assessment score compared with noncarriers (β −0.53 [95% CI, −1.02 to −0.03]). The decline was more pronounced in
DNMT3A
and
ASXL1
sequence variation carriers.
CONCLUSIONS:
In patients with AF, CHIP was independently associated with a greater burden and progression of silent brain lesions and with accelerated cognitive decline.