P1049: characteristics and outcomes of cardiovascular involvement in erdheim-chester disease

Topic: 16. Myeloproliferative neoplasms - Clinical Background: Erdheim-Chester disease (ECD) is a rare non-Langerhans cell histiocytosis. Relatively little is known about cardiac involvement in ECD, which occurs in about 40% of cases and has been associated with a worse prognosis (Gianfreda et al. 2016). Aims: To characterize ECD-related cardiac lesions and associated outcomes with respect to anatomic locations, associated cardiovascular comorbidities and prognosis. Methods: Records of patients (pts) with biopsy-proven ECD diagnosed from Jan. 1990 to Dec. 2021, seen at Mayo Clinic in MN, FL, and AZ, were reviewed. The BRAF mutation status, site and type of cardiac involvement, arrhythmias or conduction abnormalities, cardiovascular comorbidities, and first-line therapy received were captured. Cardiac involvement was diagnosed with cardiac imaging (PET-CT, cardiac CT, or cardiac MRI). All time-to-event analyses were performed using the Kaplan-Meier method. Progression-free survival (PFS) is defined from first-line therapy to first progression, not specific to ECD cardiac progression, or death. Results: Among 106 pts with ECD, 38% (n=40) had cardiac involvement. The mean age at diagnosis for the cohort was 57 years (range: 49-67); 61% were males. The median follow-up was 4.2 years (95% CI: 3.5-5.8); pts with cardiac involvement, 5 years (95% CI: 3.5-6.1), and without, 4.1 years (95% CI: 3.2-6.5), p=0.92. Sixty-five percent of ECD cardiac-involved pts (n=26) had BRAFV600E mutation compared to 42% of pts without (n=28), p=0.024. Cardiac imaging showed localization to the pericardium (n=17; 43%), myocardium (n=30; 75%), or coronary arteries (n=21; 53%). The right atrium was the most frequently affected myocardial location [RA n=29 (97%), LA n= 3 (10%), RV n=2 (6%), LV n=1 (3%)]. Infiltration around the coronary arteries was greatest around the RCA (53%) (right n=21, left n=6). Eleven (28%) of cardiac-involved pts had pericardial effusion, and 3 (8%) had cardiac tamponade. Eighteen (45%) of cardiac-involved pts had cardiac arrhythmias or conduction abnormalities, with higher rates in patients with myocardial involvement than patients without myocardial involvement (67% vs. 33%, p=0.271). Of those 12 pts, 11 had RA involvement and one had biatrial involvement. Six (33%) of those 18-pts had pericardial involvement and 8 (44%) had Infiltration around the CA. Eight pts with ECD and cardiac involvement died; one died from sudden cardiac death. Hypertension was present in 68% of pts (n=27) with cardiac involvement compared to 36% in those without (n=32), p=0.05. Current or prior smoking was present in 43% of pts (n=18) with cardiac involvement compared to 24% in those without (n=16), p=0.05. The estimated 5-year OS rate was 80% in ECD pts with cardiac involvement and 90% in ECD pts without cardiac involvement (p=0.44) (Figure 1A). The time from diagnosis to initial treatment (p=0.63) as well as the time to subsequent treatment after first-line therapy did not differ significantly (p=0.25) between ECD pts with vs without cardiac involvement. The median PFS was 2.7 years (95% CI: 2.0-8.8) in pts with, compared to 6.9 years (95% CI: 5.6-NA) in those without cardiac involvement (p=0.03) (Figure 1B). Summary/Conclusion: Our data showed that the myocardium is the most commonly affected structure in ECD with cardiac disease, and the RA is the most affected chamber. Smoking, hypertension, and BRAFV600E mutation were more common in pts with cardiac-involved ECD as compared to pts without. Although the OS in pts with cardiac-involved ECD was not different compared to those without, pts with cardiac involvement had a significantly shorter PFS compared to those without.Keywords: Histiocytosis, Hematological malignancy