Pitfalls in Morphologic Diagnosis of Pathogens: Lessons Learned From the Pseudo- Cystoisospora Epidemic

Introduction

Cystoisosporiasis is an intestinal disease of humans caused by Cystoisospora belli (formerly known as Isospora belli). It is an obligate intracellular coccidian parasite distributed worldwide, especially in the tropical and subtropical regions. Most often it causes gastrointestinal problems in immunocompromised individuals. Involvement of the gallbladder and biliary tract rarely occurs.^1-7 Within the last 5 years, multiple groups have reported parasite infestation of the gallbladder in immunocompetent patients, especially in acalculous cholecystitis.^8-17 This led to an increased screening of gallbladder specimens for the organism by the pathologist and antibiotic treatment in some of the diagnosed patients. However, recently a few groups have shown at a molecular and ultrastructural level that these are inclusions rather than actual organisms.^18-21 This review is a summary of the events which led to the development of this pseudo-epidemic, emphasizing the importance of confirmation of the morphologic findings with ancillary techniques in the accurate identification of organisms.

Beginning of the Pseudo-Cystoisospora Epidemic

Rarely, true Cystoisospora infection of the gallbladder has been reported in immunocompromised patients.^1,2,5 However, involvement of the gallbladder by these organisms in immunocompetent patients was largely unreported until few years back. Takahashi and colleagues ⁸ in 2015 were the first to report Cystoisospora in the gallbladder of an immunocompetent patient. The patient was a 47-year-old Caucasian (North American) woman with biliary dyskinesia. They described vacuolated gallbladder mucosal epithelium with eosinophilic inclusions consistent with asexual forms of Cystoisospora (Figure 1A and B). These inclusions also stained with Grocott’s methenamine silver (GMS) stain and periodic acid-Schiff (PAS) stains. No sexual forms were reported. The patient’s stool microscopic examination was negative for parasitic forms and she was treated with a course of antibiotics. The first large-series study was published by Lai and colleagues ¹⁰ in 2016. The multi-institutional study reported 18 cases of Cystoisospora infection of the gallbladder in immunocompetent individuals. Their data suggested that Cystoisospora infection was underrecognized in the gallbladders of immunocompetent patients. This was followed by multiple publications, abstract presentations at USCAP (United States and Canadian Academy of Pathology), and Twitter feeds including studies on pediatric cholecystectomies.^11-17 The majority of the studies reported an increased incidence of these organisms in acalculous cholecystectomies. Among acalculous cholecystectomies, no increased occurrence was noted in biliary dyskinesia. Our study of 1830 consecutive cholecystectomies in a recent cohort (2012-2018) and 499 in an old cohort (1997-1999) showed no significant difference in the rate of occurrence of these inclusions between cohorts. ²¹ However, none of these studies on gallbladders from immunocompetent patients confirmed the hematoxylin and eosin and special stain morphology findings with molecular or ultrastructural analysis.

OPEN IN VIEWER

Figure 1.

(A) Gallbladder mucosal epithelium with cytoplasmic vacuoles and intraepithelial inclusions giving the appearance of secretory endometrium (hematoxylin and eosin [H&E], ×100). (B) Gallbladder epithelium higher power view, showing supra- and subnuclear vacuoles with eosinophilic intraepithelial inclusions (H&E, ×400). Mimicking micro and macrogamonts. Central condensation of eosinophilic material mimicking the macrogamont. Prominent halo around mimicking parasitophorous vacuole.

Closure to the Controversy

In 2018, Swanson and colleagues ¹⁸ performed molecular analysis and proved that these structures did not amplify the internal transcribed spacer (ITS2) in the C belli ribosomal gene compared with the C belli–infected intestinal specimens from immunocompromised individuals. Their findings were reaffirmed by our study, which also utilized next-generation sequencing (NGS) of 18S RNA polymerase chain reaction (PCR) products to rule out the possibility of other parasitic organisms related or unrelated to C belli. ²¹ Further ultrastructural studies also showed this microscopic feature represents epithelial inclusions formed by aggregates of degenerated cytoplasmic structures.^19-21 All of these recent studies have suggested a possible role of bile-mediated cytolysis in the development of these inclusions. Drage and colleagues ²² reported that in their experience, these gallbladder specimens had a shorter average time to fixation when compared with the cases without inclusions. However, the mechanism of formation is still largely unknown. Recently, we detected similar epithelial inclusions in 2 routine gastric biopsies with no significant pathology. The intracytoplasmic inclusions were eosinophilic, round-oval, and oblong, resembling the macro- and microgamonts of Cystoisospora. These inclusions were mainly found in the displaced oxyntic cell aggregates in the luminal aspect of the biopsy (Figure 2). The occurrence of this inclusion in gastric biopsies seems to diminish the possible role of bile exposure in the development of these inclusions.

OPEN IN VIEWER

Figure 2.

Gastric biopsy with oxyntic cells showing eosinophilic intraepithelial inclusions resembling cystoisospora (hematoxylin and eosin [H&E], ×200).

True Cystoisospora Versus Inclusions: Compare and Contrast

In true C belli infection, multiple stages of the life cycle of the organism (sexual and asexual) are present (Figure 3A). Additionally, these forms can be seen in the subepithelial tissue as well as other tissues such as lymph nodes and spleen.^23-25 In the case reports by Benator et al ¹ and Agholi et al ⁵ on immunocompromised patients with cholecystitis, multiple stages of the life cycle were observed. ² Agholi and colleagues ⁵ also confirmed their finding with PCR-based amplification of the internal transcribed spacer (ITS1) in the C belli ribosomal gene. In contrast, in the cases reported as Cystoisospora in immunocompetent patients, the inclusion only resembled the micro- and macrogamonts of the parasite and was limited to the epithelium. These inclusions were not seen in cystic duct lymph nodes. No structures resembling oocyst or merozoites were reported with exception of Takahashi and colleagues, ⁸ who reported smaller inclusions as asexual forms. In addition, the inclusions did not elicit any inflammatory reaction. In a true intestinal infection, the organism causes villous atrophy, crypt hyperplasia, and acute and chronic inflammation with eosinophils.^23-26 Special stains (PAS, GMS, mucicarmine, and trichrome) demonstrate diffuse staining of the inclusions in contrast to the distinct punctate internal staining on PAS-D and absent staining on GMS and mucicarmine of the actual organism. ¹⁸ Ultrastructure shows the different stages of the organisms having a well-defined structure, compared with the inclusions (Figure 4A and B).^19-21 If the stool examination is not diagnostic, parasites may also be detected in duodenal aspirates or within intestinal biopsy tissue examined under light microscopy. A PCR assay has been developed for detecting Cystoisospora DNA in fecal samples. Walther and Topazian ³ have reported that PCR is a useful tool to detect Cystoisospora DNA in formalin-fixed, paraffin-embedded tissues. ³ NGS technology can also be utilized to supplement the morphologic diagnosis of organisms in formalin-fixed, paraffin-embedded tissue. ²¹

OPEN IN VIEWER

Figure 3.

(A) Common bile duct biopsy (hematoxylin-eosin [H&E] stain). (A) Inflamed mucosa with numerous intraepithelial parasites in different stages as follows: sexual form macrogametocytes (arrows) and asexual form schizonts containing developing merozoites (asterisks; H&E, ×400).

OPEN IN VIEWER

Figure 4.

(A) Cystoisospora microgametes with numerous flagellae (cilia) and condensed nuclei (center; original magnification ×9300). (B) Ultrastructure of intraepithelial inclusions (transmission electron microscopy) showing subnuclear vacuoles with electron dense structures without any specific organization (original direct magnification ×1000).

Other differential considerations for the inclusions include hyaline globules or thanatosomes. Thanatosomes are histological mimickers of these inclusions and are seen in gallbladder specimens as well as in many other benign and malignant conditions.^27,28 They are more associated with cell injury or death. In the gallbladder, they are characteristically associated with epithelial injury or inflammation, lack of distinct vacuoles, and show negative staining by Alcian blue. ¹⁷ Thanatosomes have mostly a membrane-bound ultrastructure in contrast to the non-membrane bound fibrillary aggregates seen in inclusions. ²⁷ Some groups have also suggested the possibility of abortive infection by a related animal species of the organism. ²² However, the lack of amplification of the 18S RNA PCR products on NGS analysis makes this unlikely. ²¹

Conclusion

This summary of events demonstrates the importance of supplementing morphologic diagnosis with findings from ancillary techniques, including archaic techniques like electron microscopy. In recent years, the field of microbiology has shifted significantly to molecular-based techniques due to speed and accuracy. The field of metagenomics has also grown as NGS allows shotgun approaches that characterize all organisms in a sample, as well as other features such as antibiotic susceptibility. The use of traditional and NGS-based molecular approaches to identify and subtype organisms in tissues is likely to grow immensely in the future. The C belli pseudo-epidemic highlights a situation in which earlier utilization of such molecular techniques could have prevented panic, overdiagnosis, and unnecessary antibiotic use. This experience also served as a bitter reminder to employ our critical thinking ability; that is, taking into account the associated clinical and morphologic findings (lack of symptoms and inflammation) before rendering the diagnosis of a pathogenic organism.