Long-term interval surveillance of kawasaki disease and associated coronary artery lesions using advanced coronary imaging

Introduction Kawasaki disease (KD), a systemic vasculitis with male and East-Asian preponderance, is the leading cause of paediatric acquired heart disease in developed countries. KD coronary artery lesions (CAL), ranging from ectasias (CAE) to aneurysms (CAA), not only carry significant short-term cardiac risk, but fundamentally alter vessel wall biology and accelerate atherosclerosis, even after regression (Figure). Data is sparse regarding the long-term prognosis and need for delayed surveillance using advanced coronary imaging in pre-participation settings, particularly for patients with regressed CAL and those who never had CAL, especially in multiethnic Southeast Asian populations. Purpose To assess the use of cardiac magnetic resonance imaging (CMR) and/or computed tomography angiography (CTA) for the long-term, interval coronary surveillance of patients with history of KD in a population-wide screening cohort. Methods From January 2018 to October 2023, data were collected from all conscripted young Singaporean males during a standardised pre-enlistment screening comprising health questionnaires, physical examinations and electrocardiography (ECG). KD-related medical records were obtained where available, specifically regarding the presence of previous CAL and if prior follow-up imaging showed regression. All males with history of KD were then referred for advanced coronary imaging with CMR or CTA. Although either was acceptable, CMR was generally preferred over CTA to avoid radiation. Results In our cohort of 145,000 males aged 16-22 years, 0.1% (n=138) had previous KD. Based on medical records, there were 109 (79.0%) subjects with no known CAL, 4 (2.9%) with resolved CAA, 24 (17.4%) with resolved CAE, and 1 (0.7%) with persistent CAA who was already on specialist follow-up (Table). None of the symptoms or ECG changes noted during the screening were suggestive of KD-related complications. The median time from KD diagnosis to interval CMR/CTA was 16.0 years (IQR 15.0-17.3), and most underwent CMR as opposed to CTA (91.3% vs 6.5%), with 2.2% requiring both due to suboptimal imaging on initial CMR. Amongst patients with no known CAL or regressed CAL, there were no new/recurrent CAL or stenoses detected on interval CMR/CTA. Conclusion We report important long-term data on KD derived from a rigorous, systemic screening of young Singaporean males after a median duration of 16 years from initial KD diagnosis. Based on our cohort, patients with no known CAL or regressed CAL did not newly develop or redevelop CAL, or progress to significant occlusion/calcification. Our study suggests that these patients may not need delayed surveillance with advanced coronary imaging, at least up until adolescence or early adulthood. Our results may be helpful in the future development of protocols for pre-participation screening and longitudinal follow-up of KD.