175. Clindamycin versus linezolid for Group A Streptococcal bloodstream infections?: A comparative effectiveness study of adjunctive antitoxin therapy at 137 US Hospitals

BACKGROUND: Adjunctive clindamycin has survival benefit in invasive Group A streptococcus (GAS) infections. Like clindamycin, linezolid also leads to decreased toxin and virulence factor production. However, rising clindamycin resistance among GAS isolates and inadequate clinical data on linezolid both offer pause on which to choose. We examined the impact of adjunctive clindamycin vs. linezolid on survival among patients with GAS bloodstream infections (BSI) in the presence and absence of clindamycin resistant (clinda-R) isolates METHODS: Clinical characteristics, antimicrobial susceptibility testing (AST), and antibiotic therapy were examined for unique adult inpatient encounters with GAS BSIs in the PINC-AI Database. Patients treated with a β-lactam for ≥3 days ±3 days of culture who received clindamycin ±3 days of culture were overlap weighted on a propensity-score to those who received linezolid using basine patient and hospital factors. The primary outcome was odds ratio (OR) of in-hospital mortality associated with clindamycin (vs linezolid). The secondary outcome was length of stay (LOS) among survivors. Subgroups analyses were conducted excluding clindamycin receipients with clinda-R isolates (subgroup 1) and also missing clindamycin AST(including D-test; subgroup 2). [Figure: see text] Selection of patients with Group A streptococcal blood stream infections. The database was queried for inpatients (aged ≥18 years) with blood culture displaying growth of Group A streptococcus, filtered on the basis of receiving β-lactam antibiotics within 3 days either side of culture sampling for a minimum duration of 3 days and received either adjunctive clindamycin or linezolid treatment within 3 days either side of culture sampling. [Figure: see text] RESULTS: Of 3019 β-lactam–treated inpatients with GAS BSI, 500 (17%) received clindamycin and 160 (5%) received linezolid. The prevalence of clinda-R isolates was 19%;1 isolate was linezolid resistant and excluded (Figure 1). Overlap weighting resulted in well balanced groups (Figure 2). In the overlap weighted cohort, mortality risk was similar between recipients of clindamycin (10%, [50/500]) vs linezolid (9%, [14/160]; OR 1.38 [95% CI: 0.76-2.49]). Among survivors median[interquartile range] LOS was similar between the two groups (8[9] vs. 9[8] days, p=0.45). Removing those with clinda-R isolates (N=84) and also missing AST (N=166) yielded similar results (Figure 3). [Figure: see text] Standardized mean differences for covariates included in the propensity score generation averaged across exposure categories in the unweighted cohort (blue triangles) and overlap weighted cohort weight (red circles). After overlap weighting, the mean standard difference at each variable assessed was zero . Abbreviations: ICU: intensive care unit, IVIG: intravenous immunoglobulin, NSTI: necrotizing soft tissue infection [Figure: see text] The ORs (95% CIs) of in-hospital mortality (including discharge to hospice) in the primary analysis and subgroup analysis with patients withclindamycin resistant isolates (subgroup 1) and those with clindamycin resistant isolates and missing clindamycin susceptibility results (subgroup 2) removed from clindamycin group. Abbreviations: CI: confidence interval CONCLUSION: Among β-lactam-treated patients with GAS BSI, linezolid and clindamycin displayed comparable effectiveness as adjunctive antitoxin agents. Similar intrinsic effectiveness (i.e., in patients with only susceptible isolates) supports linezolid as an alternative even in low clinda-R settings. DISCLOSURES: Ahmed Babiker, MBBS, Roche: Advisor/Consultant Morgan Walker, MD, Cytovale: Advisor/Consultant