Etiologies and Clinical Outcomes of Patients With Secondary Hemophagocytic Lymphohistiocytosis at a Tertiary PICU

OBJECTIVES: To assess the etiologies and outcomes of patients with secondary hemophagocytic lymphohistiocytosis in the PICU. DESIGN: Prospective observational cohort study. SETTING: A single PICU at a pediatric tertiary hospital in Hanoi, Vietnam. PATIENTS: Pediatric patients meeting the criteria for secondary hemophagocytic lymphohistiocytosis.None. MEASUREMENTS AND MAIN RESULTS: Between June 2017 and May 2018, 25 consecutive patients with a mean (SD) age of 23.3 months (21.6 mo) were included. Collected variables included etiologies of hemophagocytic lymphohistiocytosis and clinical and laboratory findings at admission. The Pediatric Index of Mortality 2 score at admission was calculated. Outcomes were death and multiple organ dysfunction. The severity of multiple organ dysfunction was assessed by the Pediatric Logistic Organ Dysfunction 2 score. The mean (SD) Pediatric Index of Mortality 2 predicted mortality rate was 5.6% (7.6%). Cytomegalovirus and Epstein-Barr virus coinfections (60%) were the most common suspected etiology of hemophagocytic lymphohistiocytosis. Other etiologies included Epstein-Barr virus sole infections (20%), cytomegalovirus sole infections (16%), and one unknown cause (4%). Multiple organ dysfunction (excluding hematologic failure) was found in 22 patients (88%) with death occurring in 14 patients (56%). The mean (SD) Pediatric Logistic Organ Dysfunction 2 predicted mortality rate among patients with multiple organ dysfunction was 11.9% (11.2%). Despite having lower Pediatric Index of Mortality 2 predicted mortality rates at admission, Epstein-Barr virus-cytomegalovirus coinfection cases with multiple organ dysfunction had slightly greater Pediatric Logistic Organ Dysfunction 2 predicted mortality rates than Epstein-Barr virus sole infection cases with multiple organ dysfunction: 12.2% (10.5%) versus 11.3% (11.0%). However, these rates were lower than cytomegalovirus sole infection cases with multiple organ dysfunction (14.4% [16.3%]). Area under the curve values for Pediatric Index of Mortality 2 and Pediatric Logistic Organ Dysfunction 2 were 0.74 (95% CI, 0.52-0.95) and 0.78 (95% CI, 0.52-1.00), respectively, suggesting that both scales were fair to good at predicting mortality. CONCLUSIONS: Viral infections, particularly Epstein-Barr virus-cytomegalovirus coinfections, were a common cause of secondary hemophagocytic lymphohistiocytosis. The implication of these coinfections on the clinical course of hemophagocytic lymphohistiocytosis needs to be delineated.