Diagnostic accuracy of swine echinococcosis cytopathological tests and challenges for a differential diagnosis: slaughterhouse data

Hydatidosis is of interest in many developing countries. It remains a public health concern and is on the list of neglected tropical zoonotic diseases (Moro & Shantz, 2009). Hydatidosis is caused by parasites of the genus Echinococcus (Craig et al., 2015), of which 4 are zoonotic species: Echinococcus multilocularis, Echinococcus granulosus, Echinococcus oligarthrus and Echinococcus vogeli (causing Polycystic Echinococcosis); and of two recently discovered species Echinococcus shiquicus in small mammals from the Tibetan plateau and Echinococcus felidis found in African lions (Eckert et al., 2011). The confirmation of suspected cases using adequate methods giving accurate, reliable results available on the same day of testing, is needed and essential. The tests, especially cytopathological ones, require the presence of suspected hydatid fluid. In clinical samples, many studies demonstrated the acid fast characteristics of Echinococcus spp. hook. Unfortunately, in many cases, the fluids are insufficient, no longer available, or poor around the hook. Therefore, the histopathological test is not the only remaining reference standard test (WHO/OIE, 2001; Zhang & McManus, 2006) for Cystic Echinococcosis (CE) and Alveolar Echinococcosis (AE) that demand tissue with suspected parasitic parts (Craig et al., 2015; Brunetti et al., 2010) but it is the best diagnostic method for differential diagnosis between relevant diseases such as tuberculosis, benign or malignant tumors, abscesses, or Ascaris suum infection. Moreover, histopathological tests and Polymerase Chain Reaction (PCR) accurately differentiate Echnicoccus granulosus and Echinococcus multilocularis, whereas serologic tests such as ELISA cannot (Georges et al., 2004, Thiaoying et al., 2008). Expensive PCR techniques and Western blot tests can differentiate Echnicoccus granulosus from Echinococcus multilocularis in 76 % of cases (Liance et al., 2000).

Macroscopic observation can detect Echinococcus multilocularis with multilocular cysts and smaller in size. The histopathology examination reveals the parasitic cyst on an outer laminated layer that generally calcifies and in an inner germinal layer with most often absent protoscolices. There is also a weak or no striated laminated layer of parasite membrane. Generally, vital protoscolices are absent and don’t contain calcification (Reinehr et al., 2019). The acellular laminated layer is either non-disrupted or lines the cystic parasite vesicle, which is fragmented and displays a convoluted architecture (Keutgens et al., 2013). In detail, there is noteworthy cells budding in the germinal layer; protoscolices contain calcareous corpuscles and hooks. An essential feature of vesicles is the presence of fine reticular tissue in which broad capsules contain fully developed and embedded protoscolices. Another hallmark is the presence of a germinal layer infiltrating cellular protrusion in the distant metastatic foci (Miyauchi et al., 1984, Thiaoying et al., 2008).

Macroscopic examination of Echinococcus granulosus shows all simple or unilocular cysts, much larger and fertile. Histopathologically, there is a typical trilayered cyst wall: inner germinal, intermediate laminated, and outer adventitial layer. The additional feature is the presence of vital protoscolices having calcification some-times but no calcification in the striated laminated layer (Reinehr et al., 2019). Neither brood capsule nor calcareous corpuscles are observed.

The current study aims to assess the diagnostic accuracy of two index diagnostic methods: cytopathological test 1, in which methanol is fixed followed by Ziehl Neelsen stain followed by examination under transmitted light microscopic, and cytopathological test 2, which utilizes the same stain but examination is performed with epifluorescence microscopic, and compare both with the standard histopathological method. The second goal of this study is to detect species of Echinococcus spp involved in positive swine echinococcosis, and the third objective is to demonstrate how vital differential diagnosis is.

Considering ethical issues, adverse events to animals from performing the index and reference standard tests are not applicable because they were performed as post-mortem examinations. Moreover, as in any abattoir, pigs are not slaughtered for study purposes. Therefore, formal consent is not required.

Between October and November 2019, 2524 pigs were examined by a Veterinary-Inspector for suspected echinococcosis cases, and 101 were found positive. A total of 67 suspected cases out of 101 positive samples complied with the requirements for cytopathological index tests and standard reference test examination with fluid in the cystic lesions. These 67 cases were submitted to reference histopathological test, and 63 were found positive (Fig. 1). Among these 63 were actual echinococcosis cases, there were 61 positive for Echinococcus granulosus (Table 1; Fig. 2), and 2 were positive cases for Echinococcus spp. (Table 1). Considering affected organs, most of the infestation occurred in the kidneys (55 cases; 87.3 % [95 % CI 76.5 – 94.35]), followed by the lungs (7 cases; 11.1 % [95 % CI 4.59 – 21.56]) and liver (1 case; 1.6 % [95 % CI 0.04 – 8.53]) (p < 0.001) (Table 1). All the positive histopathological cases were immuno-histochemically positive for Echinococcus granulosus. Also, the negative histopathological cases were negative for the immuno-histochemical test.

Fig. 1.

Fig. 2.

Cytopathological diagnostic tests accuracy

Methanol fixed, followed by Ziehl Neelsen staining, was examined with an epifluorescence microscope on the 67 fluids, and 47 were positive (70.1 %) (Fig. 3).

Fig. 3.

With the methanol-fixed fluids followed by the Ziehl Neelsen test, 39 out of 67 samples were positive (58.2 %). Neither index tests had intermediate or doubtful cytopathological results (Fig. 4).

Fig. 4.

Both cytopathological tests are equivalent in accuracy regarding specificity (100 % [95 % CI 100 – 100] vs. 100 % [95 % CI 100 – 100]; Mac Nemar test ratio =1[95 % CI 1 – 1]). For the PPV, cytopathological test 1 had the same accuracy as the second test (100 % [95 % CI 100 – 100] vs. 100 % [95 % CI 100 – 100], GEE Model estimate for odds ratio 1[95 % CI 1 – 1]). (Table 2).

Table 2.

Diagnostic accuracy of cytopathological test 1 (Ziehl staining with the epifluorescence microscope) and cytopathological test 2 (Ziehl Neelsen staining with the transmitted light microscope) in the detection of clinical swine echinococcosis.

	Cythopath 1% [95% CI]	Cythopath 2% [95% CI]	Test ratio[95% CI]	p value
Specificity	100 [100 – 100]	100 [100 – 100]	1 [1 – 1]	–
Sensitivity	79.66[69.39 – 89.93]	66.10 [54.02 – 78.18]	1.2 [0.88 – 1.66]	0.059
PPV	100 [100 – 100]	100 [100 – 100]	1 [1 – 1]	–
NPV	40 [18.53 – 61.47]	28.57 [11.84 – 45.3]	1.4 [0.41 – 5.2]	0.06

Examination regarding sensitivities showed that cytopathological test 1 is more sensitive than cytopathological test 2 (79.66 % [95 % CI 69.39 – 89.93] vs. 66.10 % [95 % CI 54.02 – 78.18]; Mac Nemar test ratio 1.2 [95 % CI 0.88 – 1.66]). Nevertheless, the difference in sensitivity was not significant. The NPV parameter conferred a better accuracy to the cytopathological test 1 on cytopathological test 2 (40 [95 % CI 18.53 – 61.47] vs. 28.57 [95 % CI 11.84 – 45.3]; GEE Model estimated the odds ratio 1.4 [95 % CI 0.41 – 5.2], p = 0.06). Concerning conformity of both index tests with the reference one, the cytopathological test 1 showed 82 % vs. 70.1 % for the cytopathological test 2. The Kappa parameters were 48 % [95 % CI, 22 – 75] and 34 % [95 % CI, 9 – 58]. In consequence, the corresponding Z test was Z= 3.624 (p < 0.001) for test 1 and Z = 2.69 (p < 0.01) for test 2.

Diseases of differential diagnosis detected

Many diseases have been diagnosed when swine echinococcosis is not confirmed. These diseases can be divided into parasitic, non-parasitic infections, and miscellaneous conditions such as neoplasm and cirrhosis. Indeed, diseases such as Cysticercus tenuicollis hepatitis, Ascaris suum multifocal and granulomatous hepatitis, hepatocellular carcinoma, liver cirrhosis, and other pathological conditions have been diagnosed (Table 3). Nevertheless, in one confirmed cystic echinococcosis case, it's noteworthy to mention that associated peri-portal amyloidosis affected the liver. In this challenging case, there was a yeast infection present due to Cryptococcus neoformans (Fig. 5)

Fig. 5.

Table 3.

Occurrence of diseases that could be misdiagnosed as swine echinococcosis in Côte d’Ivoire

Diseases	Involved organs
	Liver	Lung	Kidney
Ascaris suum infection or migration	4	2
Cysticercus tenuicollis hepatitis	1
Fasciolosis	1
Cryptococcus neoformans infection	1
Cirrhosis	3
Neoplasms	1(Hepatocellular carcinoma)	0	1(Small clear cell carcinoma)
Hydronephrosis			4
Actinobacillus spp or Mycoplasma spp. Pneumonia or Pleuropneumonia		11
Inflammatory conditions of unknown aetiology			2
Abscess		3
Total	11	16	7

One of the biggest challenges in meat inspection is confirming the diagnosis within hours. An accurate, less expensive and rapid test is essential for such purpose. Suppose cytopathological test results for echinococcosis diagnostic are available after several hours. In that case, it is essential to estimate their accuracy with the histopathological test as a reference standard because of the scarcity of standard gold tests. Whatever the infected organs, and whenever a cytopathological test cannot be conducted due to lack of fluid or insufficiency (< 3ml), the histopathological reference test from tissue samples has always performed well.

To the best of our knowledge, this is the first study to report the cytopathology test's diagnostic performance for the detection of swine echinococcosis. Consideration of another diagnostic test utilizing crude hydatid cyst fluid (HCF) from a specimen, such as IHA, Craig et al. (2015) reported a lower sensitivity and specificity between [25 % and 50 %] and [40 % and 60 %], respectively. For the immunoelectrophoresis test performed on the same specimen, the sensitivity estimate was 23.8 % and specificity was from 61 % to 89 % if the target population is 1 year or older. Utilizing ELISA in which recombinant antigen is based on HCF, sensitivity was 89.2 %, and specificity was 89.5 %. For many reasons, cytopathological tests’ performances are superior to these tests. First of all, with cytopathological tests, false positives are avoided. It has to keep in mind that false positive implies organ condemnation with the consequence of unjustified economic losses. This is the case with IHA, ELISA, and immunoelectrophoretic tests showing 11.5 % to 60 % false positivity. Secondly, sensitivities are similar to ELISA sensitivity or superior to immunoelectrophoresis or IHA sensitivities. Here the advantages of cytopathological tests, sensitivities, and specificities reside in the detection of true positive subjects, whereas in serological tests sensitivities and specificities are false positives due to cross-reactions. These cross-reactivities have been observed in humans infected with parasitic diseases such as cysticercosis (Hermelin et al., 2019). Checking the false negative's importance, cytopathological tests are equivalent to ELISA but superior to IHA and immunoelectrophoresis.

Concerning the applicability of index tests, they are easy to perform, results are available after several hours. In that way, the Veterinary-Inspector can rapidly provide balanced and rapid decisions. The finer point of the Veterinary-Inspector decision lies within three constraints: The necessity to preserve public health, the avoidance of unjustified organ destruction, and the need for an urgent decision to avoid keeping (in many developing countries, there is no facility to do that and if available it is very expensive) animal carcass or organs for several days pending laboratory results. Each of these aspects has double components: economic and ethical.

The major disadvantages of cytopathological tests are attributable to the disease's physiopathological state, host immune response and possible previous and intermittent treatments received by a pig. In particular, these drawbacks preferably impact the sensitivity. Firstly, there is a paucity of hydatid fluid in characteristic hooklets and protoscolices when cysts are very young. Secondly, there is an insufficient volume of hydatid fluid (< 0.3 ml) to perform the cytopathological test. The reason for that is the physiopathological standpoint; when after the host infection, the cyst grows 1mm per month (Mihmanli et al., 2016). This situation mainly occurs in earlier or more aged cysts in which water content becomes gelatinous because of its resorption and mineralization of dead protoscolices, including hooklets (Chiou et al., 2001). Third in more advanced chronic states, hydatid fluid can be very dirty due to the strong inflammatory response of the host. Consequently, this immune response causes parasite starvation (Thompson & Lymbery, 1990). In addition, many hooklets could be hidden or wrapped. Yet, it is noteworthy that the centrifugation step in our study highly overcomes this hindrance. The Kappa parameter also shows the usefulness of the two index tests. However, this is not the case for concordance coefficient, which could not be sufficient in situations when Kappa value is less than 80 %: This means there is more than 20 % of disagreement or uncertainty. So, the health and economic impacts of these incorrect results could be inacceptable for medical or veterinary practices. In all cases of hydatid fluid shortage, we recommend adding a cyst specimen to perform a simultaneous histopathological test as a guideline. This advice could be relevant in human or canine surgical treatment for echinococcosis infection.

Chemotherapy's impact on the sensitivity of index tests is primarily explained in most African regions by frequent farmers’ auto medication with drugs such as albendazole, mebendazole, and praziquantel. Indeed, these are the three most used drugs in pigs from the south region. These prescriptions could significantly lower parasitemia by rendering cysts sterile, so their mild hooklet and protoscolices contents (WHO, 2001; Arif et al., 2008). Apart from the auto medication practiced by farmers, a veterinary prescription is declined for piglets in one administration at weaning, two administrations during growth, then for sows in one administration two weeks before parturition, and all pigs in one administration whenever echinococcosis condition is suspected. This effect results in numerous false negatives with cytopathological tests. The same problem could occur in human cases, especially in developing countries where most people take drugs without medical prescription or surveillance. This would constitute a potential bias when designing a study.

Veterinary inspection as a diagnostic activity remains the basis on which the Veterinary Inspector decides on the fate of a carcass or an organ. In general, the Veterinary-Inspector can decide to accept or condemn the entire carcass or a given organ as safe for human consumption, or to accept a part of the carcass or a given organ, or to require a particular treatment of meat before approving it for human consumption. This decision-making process relies on the accurate diagnostic of the disease, its severity, and its extent in the inspected animal. Despite their sensitivities, both index tests with a PPV of 100 % mean that the probability of infection is 100 % for a given cytopathological positive test result. Therefore, Veterinary-Inspector decision-making on positive test results is reliable. Similarly, an NPV of 99 % for the two index tests implies that a negative pig tested has a 99 % chance of being uninfected. Preservation of public health due to zoonosis linked to CE, the Veterinary-Inspector decision-making has to consider additional testing such as histopathological test on suspect cyst membrane. In the west of Africa, this is the first published confirmatory diagnostic of domestic livestock echinococcosis caused by Echinococcus granulosus. The first confirmatory canine CE in west Africa was recently published (Mauti et al., 2016). This case involves genotype 6. Nevertheless, the first human cases published concern Côte d’Ivoire (Schmidt et al., 1978), Senegal (Hane et al., 1989), Ghana (Schneider et al., 2010) and Mauritania (Maillard et al., 2007; 2009). These human CE cases also involved genotype 6. It is well known that Côte d’Ivoire imported many pig breeds decades ago. Besides this, considering the use of dogs for hunting in the neighboring wild environment, the epidemiologic role mainly especially as an intermediate host of farmers’ dogs in Echinococcus multilocularis and neotropical Echinococcus species: Echinococcus oligarthrus (Lühe, 1910) and Echinococcus vogeli Rausch and Bernstein, 1972 has to be addressed in the African continent. Yet, previous studies in other countries have demonstrated the role of domestic dogs in the Echinococcus multilocularis cycle (Eckert, 1997; Gunn & Pitt, 2012), even if foxes play a major role (Kamiya et al., 1987). Moreover, the suburban location of this study, originating farms, reinforces this role exerted by farmers’ dogs. Concerning Echinococcus vogeli, the causative agent of polycystic echinococcosis, naturally infected domestic dogs have been diagnosed (D’Alessandro et al., 1981), and the bush dog is established to be a definitive natural host (Eckert et al., 2011).

Our study and many other studies show that the main locations of cystic echinococcosis involve the kidney and lungs (Chiou et al., 2001). A retrospective study between 2008 and 2015 in Côte d’Ivoire on most cattle suspect cases reported the kidneys are affected most (Acapovi et al., 2019). In Nigeria, a sole old study has shown the prevalence of suspect swine cases to be 56 % (Arene, 1985). This high prevalence could be due to many false positive cases. Nevertheless, the liver remains a potential site of infection (Moro & Shantz, 2009) even if only one case has been diagnosed in the current study. The reason for which swine kidney is preferably infected in the current study remains unknown. Further studies are needed to substantiate this renal affinity in swine.

The authors acknowledge that amyloidosis is observed for the first time in pig cystic echinococcosis, where the liver is the sole organ affected. The subtype AA, AL, or LL has not been characterized. This type of lesion has explicitly been diagnosed for Echinococcus multilocularis species infection in Cynomolgus Monkeys (Macaca fascicularis) (Bacciarini et al., 2004) and human patients (Ali-Khan & Raush, 1987). The amyloid subtype was AA, and the Disse space of the liver was affected. Additional infection, in this case owing to Cryptococcus neoformans, can be explained by pig exposure via inhalation to contaminated bird feces. Therefore, liver infection by this yeast could reveal the disease’ disseminated stage. Nevertheless, amyloidosis might be the consequence of severe and chronic CE infection.

Amongst the limitations of the current study, the histopathological reference test is to be refined with the gold standard test that rarely exists when failed to detect infected animals that have not yet developed characteristic cystic lesions. So, these animals will be tested false negative. Moreover, diagnostic accuracy between subgroups such as sex, age and breed has not been assessed yet. This aspect will be evaluated in further studies. The interval time report from the index tests and the standard reference test is not applicable because tests have been done after animals’ death. However, this interval time factor must be carefully assessed in human cases or living diseased animals because the physician or the research team often doe not get information about the drug administration.

In terms of the study added value, it should be noted the improvement in the diagnostic activity of the Veterinary Inspector who submits more or less samples to the laboratory to confirm his suspected cases

Then, based on differential diagnosis results and the macroscopic lesions generated, the college of Veterinary-Inspector brushes up its diagnostic skills before submitting samples to the laboratory. The necessity to improve disease control implies generating sound diagnostic data. It is the cornerstone condition that prevents or plans public health interventions to eradicate diseases.