Abstract 189: Natural History of Coronary Artery Aneurysms in Kawasaki Disease in US population and Risk Factors for Persistent Aneurysms

Abstract

Background: The late natural history of coronary artery aneurysms (CAA) after IVIG treatment in the US is not well described. Methods: We evaluated all KD patients (pts) at 2 centers from 1984-2014. Entry criteria were: 1) IVIG treatment; 2) CAA, defined as LAD or RCA z-score ≥ 3 or Japanese Ministry of Health criteria; and 3) ≥1 follow-up (f/u) echo. Kaplan Meier curves evaluated time to CAA regression (z < 2.5) and Cox regression examined factors associated with persistent CAA and major adverse cardiac events (MACE= death, MI, CABG, PCI, occluded CA). Results: Of 2592 KD pts, 408 (15%) met entry criteria and were 72% male; 54% white, 21% Asian, 7% black, 18% other race. Median age at fever onset was 1.8 y [IQR: 0.7-4.4y], 74% had complete KD, and fever days before 1 st IVIG were 7d [IQR 6-10d]. IVIG retreatment occurred in 35% and adjunctive anti-inflammatory therapy in 37%, both increased over time (p<.001). LAD and RCA CAA occurred in 31%, LAD alone in 47% and RCA alone in 22%. Median z-scores at CAA diagnosis were: LAD 3.61 (IQR 3.1-5.1) and RCA 3.1 (1.7-4.1) with 93 (23%) pts having giant CAA (z ≥10). Over median f/u of 2.6 y (0.01-29.0y), 313 (77%) had CAA regression at median of 1.1 mo (IQR =0.3-16.9mo). Univariate risk factors for CAA persistence were z ≥ 8 at diagnosis (HR 0.22, p <.001), earlier era (2010-2014 HR =1.0 , 2000-2009 HR = 0.60, 1990-1999 HR =0.38, 1984-1989=0.15, all p<.001), multi-vessel CAA (LCA alone HR =1, RCA alone HR =0.99 p=.99, LCA+RCA HR=0.35, p<.001) IVIG retreatment (HR 0.68, p=.003), IVIG treatment >Day 10 (HR 0.46, p<0.001), adjunctive anti-inflammatory therapy (HR 0.65, p=.03), and non-Asian race (HR 0.64, p=.001). Multivariable model for persistent CAA included earlier era (p<.001), z-score >8 at diagnosis (p<.001), multi-vessel CAA (<.001) and treatment after Day 10 (p=.01). MACE occurred in 25 (1%) pts, including 3 deaths, with univariate risk factors of CAA z-score ≥ 8 (p<.001), earlier era (p<.001), treatment after 10 d (p=.03), and non-Asian race (p=.03) Conclusion: Most CAA regress within the first year after treatment. Persistent CAA are more likely in pts with larger CAA, multi-vessel CAA and late IVIG treatment. In pts with large CAA, time to regression has gotten shorter in more recent eras, possibly related to greater use of adjunctive therapies.