739 - A Two-Center Assessment of Histopathologic Diagnostic Performance for Fungal Organism Identification

The ePoster content is available to users who are logged in.<br><br>Please log in and then come back here to use this button.

Presenting Author(s)

Cristina R. Costales

Medical Microbiology Fellow
Stanford University
Los Angeles, California
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.

Co-Author(s)

SB

Susan Butler-Wu

Associate Professor of Clinical Pathology
Director of Medical Microbiology, LAC+USC Medical Center
Los Angeles, California
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.
RS

Rosemary She

Director of Microbiology
Keck School of Medicine of the University of Southern California
Los Angeles, California
Disclosure: I do not have any relevant financial / non-financial relationships with any proprietary interests.

Background: Accurate detection and identification of invasive fungal pathogens relies on concordance of several complementary laboratory techniques, including fungal culture, serology, and histopathologic identification. Histopathologic stains such as the Gomori methenamine silver stain (GMS) are used to highlight fungal cell wall in tissue specimens. We sought to determine the diagnostic performance of histopathology fungal stains as compared to fungal culture for diagnosis of invasive fungal tissue infection at tertiary medical centers with dissimilar patient populations.

Methods: We performed a retrospective review of all surgical pathology specimens with reported GMS results and concurrent fungal culture at Keck Medical Center (Keck) and Los Angeles County + USC Medical Center (LAC). Ratios of GMS diagnostic performance were compared using chi-squared analyses, with fungal culture as the gold standard for detection.

Results: Of 1347 LAC surgical pathology specimens stained with GMS to evaluate for fungal infection, 229 (17.0%) had concurrent tissue specimens submitted for fungal culture. Of 1546 Keck GMS-stained surgical pathology specimens, 358 (23.2%) had concurrent tissue for fungal culture. GMS stain performance at LAC showed a sensitivity of 53.7% (95% CI: 37.4-69.3%) and specificity of 90.4% (95% CI: 85.2-94.2%). At Keck, GMS showed a sensitivity of 64.1% (95% CI: 52.4-74.7%), specificity of 88.9% (95% CI: 84.7-92.4%), without significant difference in performance between sites, (p=0.27) and (p=0.62), respectively. Among filamentous fungi, GMS false negative frequency at LAC was 5.3% (10/190) and 4.0% (11/277) at Keck, without significant difference (p=0.51). A subset of pathology reports suggested the fungus genus based on histologic morphology. Of 10 LAC pathology specimens with fungal genus specified, 2 (20.0%) reports gave the incorrect genus and 8/18 (44.4%) reports at Keck gave incorrect genus as per concurrent culture isolate result.Table 1. Diagnostic performance of GMS histopathology stain on surgical pathology specimens compared to tissue fungal culture at LAC and Keck Medical Centers from July 2015 through December 2018.

Conclusion: GMS stain had low-to-moderate sensitivity when compared to fungal tissue culture. Increased submission of concurrent tissue for fungal culture is likely to improve detection. When genus level identification was attempted, fungal forms were incorrectly identified in about one-third of histopathology specimens.