Revisiting acute chest syndrome (ACS) associated with sickle cell disease (SCD) vaso-occlusive pain episodes (VOE): Insight from a prospective, multicenter Phase-3 randomized controlled trial

Abstract

Abstract Background: ACS occurs in up to 20% of hospitalized patients with SCD-VOE, often prolonging and complicating hospital stay. Despite its clinical significance, variations exist in the management and outcomes of ACS across institutions. Objective: Determine prevalence of ACS in children and young adults hospitalized for SCD-VOE and describe emergency department (ED) presentation, clinical course and practice variation across institutions. Methods: Cross-sectional analysis of data collected from a PECARN-endorsed multicenter, double-blinded, randomized, placebo-controlled phase-3 trial of intravenous arginine therapy in hospitalized patients with SCD-VOE aged 3-21 (NCT04839354) at 10 pediatric EDs across the US. ACS defined by radiology-interpreted chest radiograph (CXR) positive for new infiltrate and clinical team diagnosis. ACS severity was defined a priori as mild (no oxygen (O2) use or RBC transfusion), moderate (O2 use or transfusion), or severe (bilevel positive airway pressure (BiPAP) use, intubation, or pediatric intensive care unit (PICU) transfer). Results: 271 patients enrolled (median age 15[11,18] years, 51% male; 74% HbSS/Sb°; 76% on Hydroxyurea (HU). ACS occurred in 20% (n=54; median age 13[9,16] years; 76% male; 89% HbSS/Sb°; 80% on HU); 18 diagnosed in the ED, and 36 diagnosed during their hospitalization. 72% of patients with inpatient-diagnosed ACS had a negative CXR in the ED with a mean time to diagnosis of 2.4±1.6 days. Patients with ACS at any time were significantly younger (13[9,16] vs 15[12,18] years, p=0.004), predominantly male (76% vs 45%, p<0.001), and had HbSS/Sb° (89% vs 70%, p=0.0005) vs no ACS. In the ED, ACS patients had more O2 desaturations at <94% (37% vs 12%, p<0.001), higher frequency of cough (32%vs18%, p=0.04), wheeze (15%vs6%, p=0.04), and chest pain (59%vs39%, p=0.009) vs no ACS. No significant difference in fever across groups. 83% of patients with ACS presented with a normal respiratory exam in the ED. ACS patients had lower hemoglobin (g/dL;p<0.001) and lymphocyte counts (cells/µL;p=0.02), but higher %reticulocyte (p<0.001), white blood cell counts (x109/L;p=0.001), and absolute neutrophil counts (cells/µL;p=0.02) vs no ACS. Clinical outcomes were worse in patients with ACS vs no ACS, with longer length of stay (LOS; 139[93,189] vs 71[46,113] hours, p<0.001), higher opioid utilization (2.3[0.8,3.9] vs 1.1[0.4,2.1] mg/kg, p=0.001), more transfusions (61%vs14%, p<0.001), O2 use (76%vs20%, p<0.001), BiPAP use (20%vs2%, p<0.001) and PICU transfers (15%vs1%, p<0.001). Patients with inpatient-diagnosed ACS had longer LOS vs patients with ED-diagnosed ACS (142[113,213] vs 105[52,161] hours, p=0.03). 56% of ED-diagnosed ACS was mild, with no significant difference in LOS based on severity. However, only 25% of inpatient-diagnosed ACS were mild; LOS was significantly longer in moderate/severe vs mild ACS. Viral testing was performed in 67% of ACS patients, with 36% positivity for various viral pathogens without a predominant viral isolate. Across all sites, 96% of patients received antibiotics (50% ceftriaxone, 28% ampicillin-sulbactam, 7% vancomycin, 11% other) and 82% received azithromycin. Site-level use ranged from 0-100% for ceftriaxone and 50-100% for azithromycin. 80% received albuterol, 40% received inhaled steroids, and 11% received anticoagulation. No pulmonary emboli were reported. Of the 61% of patients transfused across sites, inter-site transfusion rate varied between 0-100%. Sex, age, ED O2 saturation <94%, hemoglobin, & chest pain were independently associated with ACS in the multivariate model with an AUC=0.80. Conclusion: ACS remains common in patients with SCD-VOE (20%), with 14% of ACS diagnosed during their hospital stay. Two-thirds of subjects who developed ACS had a CXR done in the ED that was negative. ACS may have been missed in patients who did not get a CXR, and radiographic changes often lag behind symptoms. Lack of ACS-specific clinical signs and symptoms, including a normal ED respiratory exam in most patients, may delay diagnosis. Patients with inpatient-diagnosed ACS had a more severe hospital course vs those with ED-diagnosed ACS. Considerable practice variation exists in ACS management across institutions, with variable use of antibiotics and transfusion protocols. Further research on standardizing ACS treatment may be warranted. Oxygen desaturation in the ED was associated with a >3.6-fold greater risk of ACS, not included in prior ACS risk models.