Fungal Infections Associated with Contaminated Methylprednisolone Acetate in Michigan

BACKGROUND:  

On October 2, 2012, the Michigan Department of Community Health (MDCH) was notified by federal partners of a cluster of sub-acute meningitis infections suspected to be associated with preservative free methylprednisolone acetate (PF MPA) epidural steroid injections.  Three lots of PF MPA, compounded and distributed by the New England Compounding Center (NECC), were recalled and later found to be contaminated with Exserohilum rostratum, a fungus that rarely causes human disease in immuno-competent individuals.  Here we describe the epidemiology of these infections associated with the outbreak in Michigan and the impact of our coordinated public health effort.

METHODS:  

MDCH contacted all Michigan clinics in receipt of recalled PF MPA to inform them of potential risk to patients and instruct them to remove product from circulation.  MDCH worked with clinics and local health departments (LHDs) to develop potentially exposed patient lists and to begin direct patient notification.  Concurrently, MDCH worked with area hospitals and LHDs to perform active surveillance for infections associated with recalled PF MPA.

RESULTS:  

Four Michigan pain clinics received at least one recalled lot of NECC PF MPA, totaling 2,225 vials and 3,825 doses.  Approximately 2,645 of those doses were administered to 1,791 patients.  As of January 4, 2013, MDCH has identified 233 fungal infections with exposure to one of the recalled lots of PF MPA.  Of these, 26 were meningitis, 147 para-spinal infections, one stroke, 20 peripheral joint infections, 38 meningitis plus a para-spinal infection, and one para-spinal infection with a peripheral joint infection.  Eleven patients have died.  Case presentation shifted from primarily meningitis early in the investigation to para-spinal infections as time progressed.  38 patients originally treated for fungal meningitis later developed local infections at the site of their epidural injection.  To date, 70 cases have been laboratory confirmed by microscopy, histopathology, culture, and/or molecular methods.

The median age of cases is 65 years old (range: 28–89 years).  134 (58%) of cases are female.  All 233 cases were exposed at one of two Michigan pain clinics.  The attack rate is 31% at clinic A and 4% at clinic B.

CONCLUSIONS:  

The number of cases identified in Michigan increased rapidly after MDCH’s public health outreach and active surveillance.  Observations from active surveillance helped inform national clinical recommendations. Direct patient notification, instrumental in the swiftness in which patients sought diagnostics and treatment, likely contributed to better patient outcomes and fewer deaths.