577 - Incidence and Outcomes of Positive Outpatient Surveillance Blood Cultures in Hematopoietic Stem Cell Transplant (HSCT) Patients with Graft Versus Host Disease (GvHD) On High Dose ≥ 0.5 mg/kg/day (HD) and Low Dose < 0.5mg/kg/day (LD) Steroid Therapy

Background: Treatment of GvHD with steroids increases the risk of infection in HSCT patients due to additive immunosuppression and may delay the diagnosis of infection due to lack of symptoms.  Outpatient surveillance blood cultures in HSCT with GvHD being treated with HD steroids has demonstrated a blood culture positivity rate of 3.5%.  Currently, the utility of surveillance cultures in patients receiving LD steroid therapy is unknown.  Our practice includes weekly outpatient surveillance cultures for all GvHD patients treated with steroids regardless of the dose. The primary endpoint of this study was to assess the incidence of positive surveillance blood cultures in GvHD patients receiving HD or LD steroids.  Secondary endpoints included number of patients treated, hospitalization, 30 day mortality due to infection, and organisms isolated.

Methods: This was a single-center, retrospective review of GvHD patients at Yale New Haven Hospital between January 2013 and May 2019.  Patients were excluded if: lack of signs or symptoms of GvHD, treatment with steroids for any indication other than GvHD, and active GvHD without central line. Cultures from patients receiving antibiotics for concurrent infection were also excluded.

Results: A total of 71 patients met criteria with 901 blood cultures.  On HD, eight patients (14%) had 12 positive cultures (4%), and on LD, 16 patients (25%) had 22 positive cultures (4%) (p=0.15). Treatment occurred in six patients (75%) with four (24%) requiring hospitalization on HD, and 12 patients (75%) with 10 (83%) requiring hospitalization on LD (p=0.45).  The median duration of steroid therapy was 93 and 236 days with a median dose of steroids of 1mg/kg/day and 0.15mg/kg/day, respectively.  The number of positive cultures/1000 steroid days was 1.2 on HD and 0.5 on LD (RR 2.2).  30 day mortality was only noted in one patient (8%) on LD.  The most common organism in both groups was Coagulase-negative staphylococci with all six cultures on HD classified as contaminants and 6/10 cultures requiring treatment on LD.

Conclusion: Although the relative risk of positive surveillance blood cultures in HD patients compared to LD was twofold higher, there were clinically significant infections identified in the LD group.