Erythrophagocytosis is not a reproducible finding in liver biopsies, and is not associated with clinical diagnosis of hemophagocytic lymphohistiocytosis

• Published in: Virchows Archiv : an international journal of pathology 2024, Vol.484 (1), p.61-69
• Main Authors: Desai, Niyati, Kudose, Satoru, Remotti, Helen E., Del Portillo, Armando, Fazlollahi, Ladan, Lee, Michael J., Xiong, Yuqing, Moreira, Roger K., Salomao, Marcela, Fiel, Maria Isabel, Gonzalez, Raul S., Misdraji, Joseph, Gill, Ryan M., Hart, John, Kleiner, David E., Drebber, Uta, Bellizzi, Andrew M., Lagana, Stephen M.
• Format: Article
• Language: English
• Subjects: Acute Disease; Biopsy; Diagnosis; Hepatitis; Histiocytosis; Humans; Liver; Liver diseases; Lymphatic diseases; Lymphocytosis; Lymphohistiocytosis, Hemophagocytic - complications; Lymphohistiocytosis, Hemophagocytic - diagnosis; Lymphohistiocytosis, Hemophagocytic - pathology; Medical imaging; Medicine; Medicine & Public Health; Original Article; Pathology
• ISSN: 0945-6317; 1432-2307; 1432-2307
• DOI: 10.1007/s00428-023-03683-7

Hemophagocytic lymphohistiocytosis (HLH) is a rare disease with high mortality. Liver involvement is common (based on elevated liver function tests) with most patients demonstrating acute hepatitis. Liver biopsies are frequently obtained in the setting of suspected HLH for the purpose of identification of erythrophagocytosis, and if present, this finding is thought to suggest or support the diagnosis of HLH. However, there are problems with this approach; in particular, we do not know whether this finding is reproducible or whether it is specific to HLH. Therefore, we conducted a multi-institutional study in which experienced liver pathologists reviewed images taken from liver biopsies from patients with normal liver, acute hepatitis, possible HLH, and clinical HLH to determine if there was agreement about the presence or absence of erythrophagocytosis, and to ascertain whether the finding corresponds to a clinical diagnosis of HLH. Twelve liver pathologists reviewed 141 images in isolation (i.e., no clinical information or diagnosis provided). These came from 32 patients (five normal, 17 acute hepatitis, six HLH, four possible HLH). The pathologists classified each image as negative, equivocal, or positive for erythrophagocytosis. Kappa was .08 (no agreement) for case-level and 0.1 for image-level (1.4% agreement, based on two images which were universally considered negative). There was no difference in the proportion of pathologists who diagnosed erythrophagocytosis among those with different diagnoses at case or image-level ( p  = 0.82 and p  = 0.82, respectively). Thus, erythrophagocytosis is an entirely unreliable histologic parameter in liver, as it is irreproducible and not demonstrably associated with a clinical disease (namely, HLH). Unless and until more reliable guidelines can be established, pathologists should refrain from commenting on the presence or absence of erythrophagocytosis in liver biopsy.