Prevalence of cardiac sarcoidosis in white population: a case-control study: Proposal for a novel risk index based on...

Cardiac sarcoidosis (CS) is a life-threatening and underdiagnosed manifestation of the disease, which requires a complicated and expensive diagnostic pathway. There is a need for simple tool for practitioners to determine the risk of CS without access to specialized equipment.The aim of study was to determine the prevalence of CS in a group of patients diagnosed with or followed up because of sarcoidosis. A secondary objective was the search for factors associated with heart involvement.We performed a prospective case-control study (screening analysis) in consecutive sarcoidosis patients collected from October 2012 to September 2015. Cardiac magnetic resonance (CMR) imaging was performed to confirm or exclude cardiac involvement in all patients. The study was conducted in a hospital-based referral center for patients with sarcoidosis and other interstitial lung diseases.Analysis was performed in a group of 201 patients (all white) with biopsy-proven sarcoidosis, mean age 41.4 ± 10.2, 121 of them (60.2%) males. Four patients with previously recognized cardiac diseases, which make CMR imaging for CS inconclusive, were not included.Cardiac involvement was detected by CMR in 49 patients (24.4%). Factors associated with an increased risk of CS (univariate analyses) included male sex (odds ratio [OR]: 2.5; 1.21-5.16, P = 0.01), cardiac-related symptoms (OR: 3.53; 1.81-6.89, P = 0.0002), extrathoracic sarcoidosis (OR: 3.48; 1.77-6.84, P = 0.0003), elevated serum NT-proBNP (OR: 3.82; 1.55-9.42, P = 0.004), any electrocardiography abnormality (OR: 5.38; 2.48-11.67, P = 0.0001), and contemporary radiological progression sarcoidosis in the lungs (OR: 2.98; 1.52-5.84, P = 0.001). Abnormalities in echocardiography and Holter ECG were also risk factors, but not significant in multivariate analyses. A CS Risk Index was developed using a multivariate model to predict CS, achieving an accuracy of 82%, sensitivity of 50%, specificity of 94%, and likelihood ratio 8.1.CS was detected in one fourth of patients. A CS Risk Index based on the results of easily accessible tests is cost-effective and may help to identify patients who should be urgently referred for further diagnostic procedures.