Diagnosis of Inclusion Body Hepatitis in Chickens by Impression Smear Cytopathology Open Access

Aviadenoviruses are the etiologic agent of inclusion body hepatitis (IBH) and hepatitis-hydropericardium syndrome (HHS) in chickens. Emergence and reemergence of these diseases internationally and recently within the United States pose an economic and health threat to commercial flocks due to rapid, moderate to high mortality rates, poor growth, and lack of treatment options (1,2,3,4). Diagnosis of adenoviral diseases in poultry flocks represents the necessary initial step to quantifying prevalence and economic impact within regions, investigating concurrent predisposing factors, guiding selection and implementation of specific control measures, and preventing further disease spread. Currently, presumptive diagnosis of IBH can be made based on clinical signs and gross lesions; however, histopathology, PCR, or virus isolation is necessary for confirmatory diagnosis (1). These laboratory methods can be costly, laborious, time consuming, and require equipment not available in all laboratories.

Impression smear cytopathology utilizing Romanowsky staining techniques is cost-effective and technically simple, and can be performed at any size laboratory or in a field setting with minimal, commercially available reagents, contributing diagnostic data to a case within as little as 5 min (5,6). Cytopathology has been used within veterinary and human medicine for several centuries to diagnose viral, bacterial, fungal, parasitic, and other infectious and noninfectious diseases (7,8). Within poultry medicine, cytopathology has been used to characterize inflammatory patterns predictive of Marek’s disease (9) and mycoplasma infection (10), and to directly identify fowlpox and herpesvirus inclusions (11,12,13), various bacteria (14), and enteric and visceral protozoa (14,15,16,17). Despite these broad applications, the use of cytopathology in the diagnosis of aviadenoviral diseases has not yet been described. Here we evaluate the capability of Romanowsky stains to differentiate aviadenoviral inclusions and describe the cytopathologic morphology of aviadenoviral inclusions in chicken hepatocytes.

Aviadenovirus hepatitidis (also referred to as fowl adenovirus group E 8b) field isolate Ck/USA/AL/133805/LiverSpleen/2020 (133805) was propagated and titrated in chicken hepatoma (LMH, ATCC CRL-2117) cells. The titer of the stock virus was 10^5.5 TCID₅₀/ml.

Forty-one, 1-day-old, specific-pathogen-free chickens were weighed, tagged, and divided into challenge (20 birds) and control (21 birds) groups of approximately equal average body weight. Birds were orally challenged or mock challenged with 0.5 ml of 10^5.2 TCID₅₀/0.5 ml or sterile phosphate-buffered saline, respectively. Birds were housed by group in Horsfall-Bauer isolation units with access to unmedicated broiler starter feed and water ad libitum and monitored daily for well-being. Clinical morbidity was not observed during this study. At 4, 6, 10, and 14 days of age, five birds per group were euthanatized. All procedures and processes for the animal work were approved by the University of Georgia Institutional Animal Care and Use committee (AUP A2023 01-036-Y1-A0).

Between March and April 2023, 13 commercial broiler chickens from six diagnostic necropsy submissions to the University of Georgia Poultry Diagnostic and Research Center were enrolled in the study on the basis of clinical suspicion of IBH. Case histories provided by submitters included acutely increased flock mortality and uneven or reduced flock body weights in broiler flocks ranging from 25 to 36 days of age (Table 1). Birds received were dead on arrival or euthanatized via cervical dislocation at accessioning.

At necropsy, two pieces of ∼1 cm³ liver were collected from each bird, with each tissue piece designated as piece “A” or “B.” Each piece of liver was gently blotted on a paper towel, impressed sequentially along a correspondingly labeled clean glass slide (Supplemental Fig. 1), then placed within a histocassette in 10% neutral-buffered formalin. Glass slides were allowed to air dry for up to 5 min, then transported in 95% ethanol for approximately 5 min. Impression smears were stained with a Romanosky stain (Epredia Shandon Kwik-Diff, Fisher Scientific, Fair Lawn, NJ, USA) by dipping the slides through the three reagents wells of methanol, eosin-y, and methylene blue for approximately 10 sec per well. Slides were gently rinsed with tap water, allowed to air dry, and sealed using a glass coverslip and nonaqueous mounting media (OpticMount1, Mercedes Scientific, Lakewood Ranch, FL, USA). Formalin-fixed tissues were processed routinely for histopathology, and sections were stained with hematoxylin and eosin.

Per cytopathology slide and per histopathology section, 114, adequate, 400× fields (27 mm²) were evaluated by an anatomic pathologist (KM) via light microscopy and designated as “IBH-negative” or “IBH-positive” based on the absence or presence, respectively, of one or more structures compatible with an aviadenoviral inclusion body. An adequate field consisted of a single layer of at least 15 intact hepatocytes. Of 108 paired samples collected, 16 pairs were excluded because of an insufficient number of observable fields on either the cytopathologic or histopathologic component, of which three pairs were derived from clinical material, five were derived from the experimentally inoculated group, and eight were derived from the experimental control group. Morphologic changes consistent with an aviadenoviral inclusion body consisted of a hepatocyte nucleus containing an 8–15-μm-diameter, basophilic to amphophilic, globular material that disrupted to peripheralized the native chromatin and variably enlarged the nucleus by 10%–100%. Experimentally produced samples from inoculated and mock-inoculated groups were commingled, then evaluated in a blinded and randomized fashion. Clinically received samples were evaluated on a rolling basis with no blinding. Percent positive agreement and percent negative agreement between cytopathology and histopathology for all paired sample observations were determined via sensitivity and specificity calculations, respectively, with the measure of agreement determined by Fleiss Kappa analysis.

Eight, nonpathologist, veterinary professionals were provided self-study images of histopathologic IBH inclusion bodies, cytopathologic putative IBH inclusions, and cytologic normal and artifactual structures (Supplemental Fig. 2). Individuals then evaluated predetermined, randomized order, subset of 20 unknown impression smears and designated each as consistent with, or inconsistent with, IBH (Table 2). Veterinary expertise varied among individuals but included a familiarity with routine, point-of-care microscopic parasitology, with observers A, B, and C having >10 yr clinical Doctor of Veterinary Medicine (DVM) experience, observers D, E, F, and G having <2 yr clinical DVM experience, and observer H being a current DVM student. Two IBH-positive and two normal (as determined by paired histopathology) impression smear reference slides were provided alongside unknown slides. Cytopathologic designations made by observers were compared to the previously determined, paired histopathologic diagnosis per sample. Percent positive agreement and percent negative agreement were determined via sensitivity and specificity calculations, respectively, for all cytopathologic designations relative to the final histopathologic diagnosis, with the measure of agreement determined by Fleiss Kappa analysis. Positive and negative predictive values of cytologic designations were similarly calculated relative to histopathology.

The livers of experimentally infected birds rarely contained solitary to few 1–2-mm-diameter, tan to dark red foci. The livers of clinically submitted birds were variably swollen and mottled pale yellow to dark red, with disseminated, pinpoint, tan or dark red foci of necrosis and hemorrhage (Fig. 1).

Intranuclear inclusion bodies typical of aviadenoviruses (Fig. 2) were present in 14 of 35 liver histopathology sections from inoculated birds, 0 of 34 sections from mock-inoculated birds, and 19 of 23 sections from clinical birds. Putative inclusion bodies (Fig. 3A–D), based on morphologic similarity to histopathology, were observed in hepatocyte nuclei of 18 of 35 liver impression smears from inoculated birds, 0 of 34 impression smears from mock-inoculated birds, and 19 of 23 impression smears from clinical birds. Cytopathologic to histopathologic percent positive agreement and percent negative agreement for all samples were 94% and 90%, respectively, with ĸ = 0.81 (0.61–1.01, 95% confidence, P < 0.001) interpreted as “almost perfect agreement” (18). The positive predictive value of cytopathology to histopathology across all samples was 84%, and the negative predictive value was 96%. Within experimentally derived samples, percent positive agreement and percent negative agreement for all samples were 86% and 89%, respectively, with ĸ = 0.60 (0.36–0.84, 95% confidence, P < 0.001), with a positive predictive value of 67% and a negative predictive value of 96%. Within clinically derived samples, percent positive and negative agreement and positive and negative predictive values were all 100%, and ĸ = 1 (0.59–1.41, 95% confidence, P < 0.001). When present, the total inclusion body count ranged from 2 to 50 (median = 8) histologically and 1 to 46 (median = 4) cytologically per experimentally derived sample and from 58 to 6402 (median = 201) histologically and 11 to 504 (median = 44) cytologically per clinically derived sample. Bacterial hepatitis was diagnosed histopathologically in 4/23 clinical liver sections that lacked inclusion bodies.

Observers collectively averaged 66% positive agreement and 98% negative agreement to the histopathologic diagnosis (Table 2), with ĸ = 0.61 (0.53–0.68, 95% confidence, P < 0.001) interpreted as “substantial agreement” (18), yielding a positive predictive value of 99% and a negative predictive value of 67%. Interobserver agreement was higher (76% positive agreement, 94% negative agreement, ĸ = 0.68 [0.65–0.71], 95% confidence, P < 0.001) within the 12 of these samples that originated from natural disease (Table 2).

IBH is a prevalent disease within commercial broilers that can cause moderate to high mortality within flocks, and diagnosis currently relies on laboratory methods that can be costly in terms of time, money, and labor (1). The results of this study support Romanowsky-stained impression smear cytopathology of the liver as a useful screening technique for inclusion body hepatitis in chickens. Screening tests can offer advantages within a health care program via increased accessibility and decreased time and cost; however, these efficiencies may come with the disadvantage of utilizing a less accurate test (19). Thus, screening tests are employed to distill diagnostic options such that the decision-making process is guided towards confirmatory tests of high relevance and accuracy, and resources are not wasted on extraneous efforts (19). Impression smear cytopathology represents a highly accessible and portable means of tentative diagnosis of IBH at the time of necropsy for veterinarians in locations without immediate access to a diagnostic laboratory. This rapid and cost-effective diagnostic screening option increases the opportunity to identify and track positive cases, communicate to affected parties in a timely manner, and focus valuable time and monetary resources on specific confirmatory laboratory tests or response measures.

When evaluated by a pathologist, IBH cytopathologic diagnosis was that of perfect agreement to histopathology among clinically derived samples, and “almost perfect agreement” to histopathology overall. Within experimentally derived samples, the rarity of inclusion bodies both cytologically and histologically likely contributed to the lower agreement and predictive values associated with pathologist evaluation. In several of these sample sets, IBH presence was determined by one or two inclusions total on either the cytopathology or histopathology component, resulting in discrepant diagnoses if no inclusions were also located within the paired sample. These results highlight a limitation of both cytopathology and histopathology in the definitive diagnosis of IBH, because both modalities are dependent on sample collection coinciding with a window of high viral replication with inclusion body formation. As these mismatched diagnoses occurred only within the experimentally inoculated group, and no incongruous designations occurred within the mock inoculated group or clinically submitted cases, these results support pathologist-evaluated cytopathology as a comparable method to histopathology in cases of natural IBH infection exhibiting clinical morbidity.

When evaluated by nonpathologist veterinarians, IBH cytopathologic diagnoses maintained “substantial agreement” to histopathology. Notably, the high positive predictive value highlights the rarity of false positive calls made within the veterinary observations group, regardless of personal experience level. Here this high degree of specificity invokes the opportunity to “SPIN,” in which a “Specific test when Positive rules IN the disease” (20,21).

As positive predictive value increases with high disease prevalence (20), the practicing veterinarian is reminded to account for local disease pressure when encountering a suspected cytopathologic IBH inclusion body in clinical cases, and to modulate their degree of diagnostic confidence accordingly. In contrast, the practitioner is cautioned against employing “SNOUT,” in which a “Sensitive test when Negative rules OUT the disease” (20,21), due to the relatively lower sensitivity and negative predictive value achieved by nonpathologist-evaluated cytopathology in this study. Interestingly, a higher degree of positive agreement (“sensitivity”) was present among observers for clinically derived samples than for experimentally derived material. This is suspected to be because of the higher density of inclusion bodies present in these clinical samples, offering more opportunities to recognize the pathologic feature. However, in these clinical samples, disparity of inclusion body density was noted between cytopathology and histopathology, often observed as extensive areas of cellular and nuclear rupture on impression smear. This cell rupture was attributed to the increased fragility of diseased hepatocytes compared to their normal counterparts, emphasizing the importance of obtaining samples of high diagnostic quality using gentle impression technique during specimen preparation. In conclusion, the nonpathologist veterinarian is encouraged to utilize this testing modality as a screening function for IBH and is encouraged to pursue further confirmatory diagnostic testing as guided not only by the cytopathology results, but also by their comprehensive assessment of all factors contributing to disease presentation within a flock. Further sensitivity and specificity characterizations of IBH-targeted cytopathology could be investigated using a larger sample set, sampling of inoculated birds exhibiting morbidity, blinded evaluation of clinically derived samples, and comparison of cytopathology to other laboratory tests, such as PCR or virus isolation.

Supplemental data associated with this article can be found at 10.1637/aviandiseases-D-24-00080.s1.

Supplemental Fig. 1. Impression smear technique. (A) Fresh liver is gently blotted until slightly tacky. (B) Liver is gently impressed along a clean glass slide. (C) Insufficient blotting results in areas that are too thick (arrow) to evaluate. (D) Proper blotting allows a single layer of cells across the majority of the impression smear.

Supplemental Fig. 2. Nondiagnostic structures that should not be interpreted as inclusion bodies. (A) Large and small lymphocytes (asterisks) (B) Monocytes (arrowheads). (C) Overlapped nuclei (circled). (D) Ruptured nuclei (brackets).

The authors thank Susan M. Williams, Karen Grogan, Harlee Ingram, Llewelyn Sellers, Brittany Buckley, Will Byrum, Erich Linnemann, and Angel Olvera for their assistance in case management and sample processing.

DVM =

Doctor of Veterinary Medicine;

HHS =

hepatitis-hydropericardium syndrome;

IBH =

inclusion body hepatitis;

LMH =

chicken hepatoma cells