Abstract Background and Aims Infection is the leading cause of death in elderly patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV). Older patients are more frail and have poorer nutritional reserves. The decrease in skeletal muscle mass and the development of sarcopenia is a pervasive problem among the elderly. Computer tomography (CT) scan provides a practical method to measure muscle mass in hospitalized patients which has been shown to predict the mortality in cancer and ICU patients. In the present study, we aim in to investigate the relationship between skeletal muscle area and clinical outcomes in elderly AAV patients with renal involvement. Method Hospitalized AAV patients older than 65 years at the diagnosis in Peking Union Medical Hospital from June, 2014 to June, 2019 were enrolled. Skeletal muscle area was calculated on CT image at the third lumbar vertebra level with previously published method using NIH ImageJ software. Baseline Birmingham Vasculitis Activity Score (BVAS), lab results, intensity of immunosuppressive treatment, follow up time and status of the endpoints (infection or death) were recorded. Characteristics between patients with or without endpoint event were compared. Univariate and multivariate Cox regression model was used to determine the independent predictors of mortality. Results A total of 58 patients were included, with. 48.3% male (n=28) and mean age 71.11(65.13, 78.13) years. Baseline BVAS score was 17±4.36 and eGFR was 18.67 (8.306, 34.59) ml/min×1.73m2. Muscle area was 119.6±24.14 cm2 measured on the first CT scan after admission. The baseline serum albumin was 34 (29, 37) g/L, baseline Hemoglobin was 97.83±21.81g/L and baseline immunoglobulin G level was 14.18±4.83g/L. Treatment include glucocorticoid (100%), cyclophosphamide (87.9%), pulse glucocoritcoid (36.2%) and plasma exchange(25.9%). After a median follow-up of 233.5(56.75, 435.8) days, 30 patients experienced 36 episodes of infections that required either hospitalization or administration of antimicrobial agents, of which 7 episodes were complex infections of more than one site and 10 were infections resulting in ICU admission. Median time to infection was 16 (7.5, 59) days from admission. Pneumonia (n=30) and CMV viremia (n=7) were most common. Seven patients died of infection and 1 died of gastrointestinal bleeding. Median time to death was 44 (25.5,75.3) days from admission. In univariate analysis, only age (HR=1.198, CI 1.046,1.372, p=0.006) and muscle area (HR=0.945, CI 0.907,0.986, p=0.009) significantly predicted death. Other variables including BVAS score, eGFR at the onset of disease, history of diabetes, history of pulmonary disease, disease affecting the respiratory system and whether receiving glucocorticoid pulse therapy and baseline nutritional markers i.e. BMI, serum albumin level, lymphocyte count, hemoglobin level and immunoglobulin G level were not significant predictors of death or infection. In multivariate survival analysis, muscle area were significant predictor of 6 month death even when controlled for baseline BMI and eGFR. Conclusion Muscle area as measured at L3 vertebrae level on non-contrast enhanced CT is an strong predictor of 6 month mortality in elderly AAV patients receiving immunosuppressive therapy
Read full abstract
Sep 1, 2024
Junjie Chen + 6
Open Access
