383. An Increased Rate of Candida parapsilosis Infective Endocarditis Is Associated With Injection Drug Use

Abstract

Background. Candida parapsilosis fungemia typically occurs in patients with intravascular catheters or prosthetic devices. In 2017, we noted an increase in C. par-apsilosis infective endocarditis (IE). Methods. We retrospectively reviewed C. parapsilosis fungemia and IE from January 2015 to February 2018. Species were identifed using MALDI-TOF, and con-frmed by ITS sequencing. Results. Between 2010 and 2017, there was no increase in cases of C. parapsilosis fungemia (mean: 13/year), but there was a signifcant increase in C. parapsilosis IE (P = 0.048) (Figure 1). From January 2015 to February 2018, 22% (12/54) of C. parapsilosis fungemia was complicated by IE. Demographics of C. parapsilosis fungemia included: community-acquired infection (87%), presence of vascular catheters (80%), opiate noninjection drug use (non-IDU, 44%), IDU (20%), and presence of cardiac devices (18%). Ninety-one percent (49/54) of C. parapsilosis fungemia was caused by C. par-apsilosis sensu strictu (Cpss); C. orthopsilosis and C. metapsilosis accounted for 4% (2/54) each (1 isolate could not be subtyped). Cpss, C. orthopsilosis, and C. metapsilosis accounted for 83% (10/12), 8% (1/12), and 8% (1/12) of IE, respectively. Ninety-two% (11/12) of C. parapsilosis IE was lef-sided, and 33% (4/12) involved multiple valves. Risk factors for C. parapsilosis IE were past or active IDU (P < 0.001), community-acquired fungemia (P = 0.02), prosthetic heart valve (P = 0.01) or implanted cardiac device (P = 0.03). Receipt of an antibiotic within 30 days was a risk for C. parapsilosis fungemia without IE (P = 0.001). Median age for IE vs. fungemia was 38 vs. 57 years (P = 0.09). By multivariate logistic regression, IDU (P < 0.0001), prosthetic valve (P = 0.006) or implanted cardiac device (P = 0.04) were independent risks for C. parapsilosis IE. 70% (7/10), 20% (2/10), and 10% (1/10) of patients with IDU and C. parapsilosis IE primarily used heroin, buprenorphine/naltrexone, and cocaine, respectively. 50% (6/12) of patients with C. parapsilosis IE underwent surgery; most common initial AF regimens were caspofungin and amphotericin B. Nonsurgical patients were suppressed with long-term azole; one relapsed requiring surgery. Tirty-day and in-hospital mortality for patients with fungemia vs. IE were 32% vs. 17% and 26% vs. 17%, respectively. Conclusion. C. parapsilosis IE has emerged at our center. Unique aspects of C. parapsilosis pathogenesis that may account for emergence are a propensity to colonize skin, adhere to prosthetic material and form bioflm. C. parapsilosis IE may be an under-appreciated consequence of IDU and opioid abuse.