Optimizing routine screening for cardiac sarcoidosis through use of commonly available studies

Sarcoidosis is a multisystem granulomatous disorder with unclear etiology. Morbidity and mortality vary based on organ involvement, with cardiac sarcoidosis (CS) associated with higher mortality; despite this, CS remains underdiagnosed. The Heart Rhythm Society (HRS) expert consensus statement recommends screening sarcoidosis patients for CS utilizing a symptom screen, EKG, and echocardiogram (TTE), while the American Thoracic Society (ATS) guideline recommends only EKG and symptom screening. These recommendations, however, are based on limited data with recommendations for further studies.

The purpose is to evaluate the prevalence of abnormal screening tests in patients with sarcoidosis and the correlation of these tests with the subsequent diagnosis of CS. A specific emphasis was placed on evaluating the sensitivity of the recommendations versus the sensitivity of a modified criteria.

and Methods: This study retrospectively evaluated a database of prospectively enrolled patients from a tertiary military academic center. All patients who underwent imaging with cardiac MRI and/or FDG-PET were identified. These results were correlated with screening studies (symptom screen, EKG, TTE, and ambulatory rhythm monitoring (ARM)) and used to calculate sensitivity, specificity, and positive and negative predictive values for each test. Using a clinical diagnosis of CS as the reference standard, the sensitivity and specificity of the HRS criteria were calculated and compared to a modified screening rubric developed a priori, consisting of minor changes to the criteria and the addition of ARM.

This study evaluated 114 patients with sarcoidosis with 132 advanced imaging events, leading to a diagnosis of CS in 36 patients. Utilizing HRS screening recommendations, the sensitivity for CS was 63.9%, while the modified criteria increased sensitivity to 94.4%.

This study suggests that the HRS guidelines lack sensitivity to effectively screen for CS and that a modified screening model which includes ARM may be more effective.