Introduction

Toxocariasis in humans occurs by ingestion of embryonated Toxocara canis or Toxocara cati eggs. The hatched larvae make their way to the intestine, and subsequently penetrate the intestinal wall and migrate to the liver, where most of the parasites are sequestered in an arrested state of maturation. From the liver, some parasites may migrate to the lungs, muscles, eyes, or central nervous system. The presence of Toxocara larvae in tissues results in a wide spectrum of clinical illness, from visceral larva migrans or visceral toxocariasis (VT) [1], to ocular larva migrans or ocular toxocariasis (OT) [2], to neural larva migrans or neurotoxocariasis (NT) [3]. In addition, covert and common toxocariasis, which are predominantly seen in children and adults, respectively, are recognized as asymptomatic forms of infection [4, 5].

Establishing a diagnosis of toxocariasis in humans is difficult, largely because the clinical syndromes can be vague, and general laboratory findings may be associated with other parasitic and viral infections and immune system abnormalities. Toxocara do not develop to the adult stage in the human host, and therefore eggs are not shed in feces. As such, classical ova and parasite examinations are not useful for diagnosis of human toxocariasis. In most cases of toxocariasis, small numbers of tissue-dwelling larvae are widely distributed throughout the body. A definitive diagnosis may be made in rare instances when Toxocara larvae are confirmed by histological examination of biopsies of lesions in affected tissue, but biopsies are typically not performed merely for diagnostic purposes, as the probability of obtaining tissue containing a Toxocara larva is low. In some patients with ocular infection, viable larvae may be seen with the use of ophthalmoscopy; however, these cases are not the norm. Because of these limitations in observing and identifying Toxocara, diagnosis of toxocariasis is almost always accomplished using immunodiagnostic methods.

Parasitological diagnosis

Although rare, parasitological identification of Toxocara is possible, especially if larvae are recovered from ocular fluid or cerebrospinal fluid (CSF). Larvae recovered in these specimens are typically viable and are very motile. The arrested Toxocara canis larvae found in humans are approximately 400 μm in length and 18–20 μm in diameter in the mid-region. Toxocara cati larvae are slightly thinner, about 16 μm in diameter [2]. T. canis has spear-shaped cervical alae, while the alae of T. cati are described as arrow-shaped [6–8]. Histological examination of tissues may reveal partial or intact larvae [1].

Immunodiagnosis

Early serology tests utilized extracts prepared from adult T. canis worms and contained epitopes that were highly cross-reactive with Ascaris and other helminths. Research has shown that adult worm preparations are not useful for immunodiagnosis because most of the larval antigens recognized by infected human hosts are not shared with adult worms [9]. In 1975, de Savigny described an in vitro culture system that supported maintenance of T. canis larvae in serum-free media [10]. In vitro cultured larvae are the ideal source of material for immunodiagnosis, as the larvae remain motile and metabolically active and secrete antigens for months, ostensibly mimicking human toxocariasis. A detailed procedure for generating the Toxocara excretory-secretory antigens (TES-Ag) is presented in a recent review by Fillaux and Magnaval [11]. The availability of TES-Ag led to improved diagnostics for toxocariasis, and in 1979 the TES-Ag was used to develop an ELISA [12]. The TES-Ag ELISA, with a sensitivity of 78 % and a specificity of 92 % [13], represented a significant improvement in diagnostics for toxocariasis, although cross-reactions with other nematode species have been reported in the years since the original report [14–16]. Overall, assay performance has remained consistent over many decades, with variations depending on the cutoffs used. CDC reports a sensitivity of approximately 75 % in patients with clinically diagnosed VT and a specificity greater than 90 % at titers of ≥1:32 (personal communication) for the TES-Ag ELISA. The presence of cross-reactive antibodies is especially problematic in tropical regions, where infection with multiple helminths is common [17, 18]. Some have proposed absorbing the cross-reacting antibodies using antigen extracts from related nematodes, particularly Ascaris, prior to testing in the TES-Ag ELISA as a means to improve specificity [17, 19]. TES-Ag ELISA results are often reported as a titer, although results from commercial kits can be semiquantitative or qualitative (positive or negative). As a rule, serology cannot differentiate current from past infection.

Confirmation of positive TES-Ag ELISA results using the TES-Ag immunoblot (IB) improves specificity [20, 21]. Antibody reactivity to the lower-Mr TES antigens (24, 28, 30, and 35 kDa) is more frequently associated with toxocariasis than reactivity to higher-Mr antigens (50, 81, 132, 147, and 200 kDa). Reactivity to the higher Mr antigens is associated with infections caused by other helminths [21–24].

Detection of Toxocara-specific Ig subclasses and IgG isotypes

The classical TES-Ag ELISA detects Toxocara-specific IgG. Detection of specific IgE has been shown to improve the specificity of the TES-Ag ELISA and the TES-Ag IB [25, 26], although others have suggested that specific IgE detection is less sensitive than IgG detection [19, 27]. In France, an IgE TES-Ag ELISA is routinely used for clinical diagnosis [25]. Specific IgM is not useful in diagnosing acute infection, however, as Ig antibodies persist throughout the course of infection [27], likely because Toxocara-specific glycans on the surface of the parasite continue to induce parasite-specific IgM [28].

TES-Ag-specific IgG isotypes have been evaluated as possible options for improving the sensitivity and specificity of the TES-Ag ELISA. Research has shown that all four human Ig isotypes (IgG1–4) are produced to the TES-Ag. In one study, specific IgG2 was shown to be the most sensitive and IgG3 was shown to be the most specific and accurate, although the sensitivity of total specific IgG was only 50 % in this study [18]. Others have shown that detection of specific IgG4 can increase specificity, especially if used as a confirmatory test for positive TES-Ag IgG ELISA results [18, 29, 30].

Antibody persistence

Arrested larvae are long-lived, surviving in human tissue for months to years and continuing to release antigens that serve as ongoing stimulus for immune response. The length of time that specific antibodies persist after therapeutic cure or natural resolution infection is unknown [31]. As the risk of infection is lower in adults compared to young children, it is not surprising that studies comparing antibody titers in preschool children to adults showed that antibody levels decreased with age [32], which suggests that antibodies gradually decrease after the parasite dies. One study that monitored antibody levels after anthelminthic treatment showed that specific IgG levels eventually decline but remain detectable for 4–5 years. IgE levels decreased but were still measureable one year following treatment [19, 33••]. Other studies reported that antibodies to the TES-Ag, when measured in the TES-Ag IB, persisted for about 1 year [11]. Evaluations of tests that measure the avidity of specific IgG have shown that they are able to differentiate newer from older infections, although this method has not been adopted to date for clinical diagnostic purposes [27, 34].

Antigen detection

Because the arrested Toxocara larvae remain metabolically active and produce substantial amounts of antigen [10], tests to detect circulating Toxocara antigens would seem to be the method of choice for differentiating active versus resolved infections. Monoclonal antibodies generated against the TES-Ag have been incorporated into antigen capture assays, with mixed results. Some monoclonal antibodies appear to be species-specific, while others detect both T. canis and T. cati antigens [35–37, 38••]. While some authors report detection limits as low as 5 ng/mL, most studies show that patients with inactive disease are less likely to have detectable serum concentrations of circulating antigen [36]. Questions also remain as to the specificity of some of the monoclonal antibodies used to date, as up to 25 % of sera from patients with other helminth infections also demonstrate reactivity. It is unclear whether this reactivity is from Toxocara antigens due to latent undiagnosed toxocariasis in these patients or from the presence of shared antigens between helminths.

Molecular detection

Molecular methods using serum have not proven useful for the general clinical diagnosis of toxocariasis. However, application of PCR-based tools may be of value in the detection of Toxocara DNA in human biopsy specimens. PCR assays that amplify sequences of the internal transcribed spacer (ITS) regions ITS-1 and ITS-2 of ribosomal DNA, which can differentiate T. canis and T. cati, have been used to identify Toxocara in animal tissues, bronchial alveolar lavage, and ocular fluids [39, 40•, 41, 42]. In one study, PCR was used to detect parasite DNA in human OT, and although no positive specimens were detected in this study [43], the application of PCR for diagnosis of OT seems reasonable.

Immunodiagnosis of different forms of toxocariasis

Visceral toxocariasis

Visceral toxocariasis occurs when Toxocara larvae migrate to the liver and other organs. The severity of symptoms in human toxocariasis is proportional to the number of larvae ingested [44] as well as trapping of larvae in the liver [45]. It is not surprising, therefore, that VT was originally described in children and associated with children who had a history of geophagia. Characteristic manifestations of VT include fever, cough, wheezing, abdominal pain, and hepatomegaly, and in rare instances, pneumonia, myocarditis, and encephalitis. General laboratory findings include leukocytosis, hypereosinophilia, hypergammaglobulinemia, and elevated serum IgE. Asymptomatic or inapparent infection with Toxocara, with or without eosinophilia, was also recognized in children [4]. A visceral larva migrans syndrome was later described in adults, which was characterized by symptoms and signs consistent with allergic response: pruritis, rash, urticaria, respiratory problems, cough, eosinophilia, and elevated IgE levels. Hepatomegaly was observed in these patients as well [5]. Since clinical and laboratory findings may be nonspecific, positive serology results from the TES-Ag ELISA are important in establishing a diagnosis of VT [46].

Not only the severity of disease, but the level of antibody response induced in VT is proportional to the infective dose as well [45]. The utility of the TES-Ag ELISA for diagnosis of VT was originally evaluated using 62 sera from cases with presumed clinical VT, which was defined as the presence of 5 of the 6 following criteria: 1) leukocytosis (white blood cell count >10,000/mm3), 2) eosinophilia (>10 %), 3) elevated anti-blood group A isohemagglutinins, 4) elevated anti-blood group B isohemagglutinins, 5) hypergammaglobulinemia; and 6) hepatomegaly [13]. Using these specimens, the sensitivity and specificity of the TES-Ag ELISA were determined to be 78 % and 92 %, respectively. In the study that originally described covert toxocariasis in children, the concentration of specific antibody was higher in children with eosinophilia [4], confirming studies in animal models showing that symptom severity and antibody were proportional to infective dose [44].

Ocular toxocariasis

The occurrence of ocular invasion is almost always unilateral and results in endophthalmitis or retinal granulomas. Signs of VT are usually not present, and notably, eosinophilia is typically absent. Diagnosis is primarily clinical; larvae are demonstrated only after histopathologic evaluation of an enucleated eye. As with VT, serological testing is useful for diagnosing OT. However, the TES-Ag ELISA is not as sensitive for diagnosis of OT as it is for VT. In one case–control study that evaluated sera from 17 defined cases of OT using the TES-Ag ELISA, only 65 % of cases were seropositive. Sera collected within the first month of infection were more likely to be positive, although some sera were positive even when collected years after infection [47].

Measurement of intraocular antibodies has been shown to improve the sensitivity of OT diagnosis [48–50]. One study compared intraocular antibody levels to serum levels in 49 patients, (37 adults and 12 children) with undetermined posterior uveitis. Five adults were seropositive, but none produced intraocular antibody. In contrast, one-fourth of the children had detectable intraocular IgG but only one child was seropositive. All had higher antibody levels in the vitreous humor than in serum [50]. In another study, serum and aqueous fluid were tested in 14 patients with suspected OT; 12 serum and 12 aqueous humor specimens were positive [48]. Interestingly, 33 % had peripheral blood eosinophilia and 43 % had elevated serum IgE, suggesting that OT was strongly suspected in this group of patients. Ocular fluids typically do not contain an abundance of nonspecific antibodies, so testing can be performed on undiluted or minimally diluted eye fluids, although this has not improved sensitivity [47, 51]. In the clinical setting, testing for intraocular antibody in aqueous or vitreous humor is recommended over serum antibody testing if OT is suspected, but both specimen types should be submitted. Serology is useful for diagnosing OT if TES-Ag results are positive, but OT cannot be ruled out when results are negative.

Neurotoxocariasis

Neurotoxocariasis (NT) is a rare manifestation of toxocariasis characterized by acute eosinophilic meningitis. Serology is almost always necessary to confirm diagnosis, although Toxocara larvae may be recovered in the CSF in rare instances. Detection of antibodies in CSF has been shown to be more sensitive than testing serum in patients with NT [52, 53]. In some instances, particularly if the infective dose is high, symptoms may precede antibody development. If a negative serology result is obtained, an additional specimen should be collected one week later and tested. If ocular larva migrans is also suspected, aqueous and vitreous humor should be examined.

When NT is suspected, serology for baylisascariasis should be performed in parallel with serology for toxocariasis to rule out infection with Baylisascaris. Patients infected with Toxocara produce antibodies that cross-react with Baylisascaris procyonis ES antigens (BpES-Ag) (Table 1). This cross-reactivity is seen in both the BpES-Ag ELISA and BpES-Ag IB tests. In general, serum antibodies from the limited number of patients with Baylisascaris infections do not cross-react with TES-Ag [54, 55], although one case of baylisascariasis has been reported with high titers to both Toxocara and Baylisascaris antigens [56]. A patient with positive serology results in both toxocariasis and baylisascariasis tests is usually presumed to have toxocariasis [57]. A diagnosis of baylisascariasis can be established if a patient’s serological testing results are negative for toxocariasis and positive for baylisascariasis. Importantly, in recently developed serological tests for baylisascariasis that utilize recombinant protein antigens, antibodies generated in patients with toxocariasis do not cross-react [58•].

Table 1 Reactivity of serum and CSF antibodies produced in NLM syndromes in tests for toxocariasis and baylisascariasis reaction in patients with
Full size table

Best practices for diagnosis of toxocariasis

The currently accepted best practice for immunodiagnosis of clinical toxocariasis is to detect Toxocara-specific IgG using the TES-Ag ELISA. In Europe, TES-Ag ELISA positive results are confirmed using a TES-Ag IB [21, 25]. Testing in the U.S. is limited to TES-Ag ELISA, and IB is not performed. Commercial TES-Ag ELISA kits available in the U.S. and Europe report improved sensitivities and specificities of approximately 90 % [59, 60]. An IB test that utilizes TES-Ag is manufactured and sold in Europe but cannot be imported into the U.S. for clinical diagnosis. If OT is suspected, serum and vitreous or aqueous humor should both be submitted. Likewise, in NT, both serum and CSF should be tested, and baylisascariasis should be ruled out. A clinically positive result in an asymptomatic patient does not have diagnostic value and should be considered only when all other etiologies have been ruled out [11].

Opportunities for improved immunodiagnosis

Because the TES-Ag ELISA is essentially the sole test used for clinical diagnosis and epidemiological studies, some have suggested that the test is standardized and validated [27]. While it is true that a relatively similar antigen preparation is utilized by virtually all clinical microbiologists and researchers, there is no globally accepted method for establishing a cutoff or even for defining reportable units (titers, ratios, absorbance, or optical density units). Lot-to-lot variation in the TES-Ag preparations is also a recognized drawback. Importantly, the TES-Ag ELISA and IB are dependent on in vitro cultured Toxocara larvae to maintain the supply of TES-Ag. In addition to the aforementioned cross-reactivity with other helminths, other limitations include reduced sensitivity (approximately 50 %) for detecting OT, reduced sensitivity for detecting low-level infection, possible reduced sensitivity for detecting T. cati infection [61], and lengthy antibody persistence, lasting up to 5 years after treatment. These shortcomings are driving research to identify better diagnostic tools for human toxocariasis.

The development of novel tests utilizing recombinant antigens to detect specific antibodies represents the single greatest opportunity for improved diagnostics of human toxocariasis. Much of the information that can inform the identification of new targets and the development of improved immunodiagnostic tests has come from previous work on the characterization of the TES-Ag. The TES-Ag comprises approximately 10 major proteins based on polyacrylamide gel electrophoresis and approximately 50 proteins when characterized by 2D gel electrophoresis [30]. Histochemical studies and later sequence analysis have suggested that TES-Ag molecules are glycoprotein in nature and are heavily glycosylated, with up to 40 % of the mass attributable to carbohydrate [62]. The TES-Ag IB has helped identify antigens with diagnostic relevance. The lower-molecular-weight TES antigens (24, 28, 30, and 35 kDa) appear to be better candidates for immunodiagnosis than the larger antigens, as the higher-molecular-weight antigens are associated with cross-reactivity [17, 21–24].

The genes for many of the proteins identified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) have been cloned and expressed in bacterial or eukaryotic expression systems (reviewed in [63]). The TES-Ag includes mucins, C-type lectins, a phosphatidylethanolamine-binding protein, and a number of enzymatic proteins.

Several of the recombinant forms of the TES-Ag have been analyzed for diagnostic utility (Table 2) [60, 64••, 65–68]. Although these studies are limited in number and scope, findings suggest that recombinant proteins can be used successfully for sensitive and specific immunodiagnosis of toxocariasis. Interestingly, although the native TES-120 is associated with cross-reactivity, when the recombinant protein is expressed in E. coli or in Pichia pastoris, it appears to perform well as a specific diagnostic antigen [64••, 65, 66]. In another study, a combination of three recombinant proteins (rTES-26, rTES-30, and rTES-120) proved to be more sensitive than any single recombinant protein alone [64••]. When the performance of the three recombinant proteins was compared, the antigen that represented the lowest Mr native protein (TES-26) had the lowest sensitivity and the highest specificity, which is the inverse of what is seen with the native proteins. Additional studies are needed to confirm these initial observations in order to determine if or when recombinant proteins can replace the TES-Ag for clinical diagnosis.

Table 2 Proteins with potential for immunodiagnosis of toxocariasis
Full size table

Research gaps

In the development of new immunodiagnostic tools, several particulars should be addressed. In addition to determining sensitivity and specificity in more populations with more samples, evaluations of new recombinant protein-based assays must also assess suitability for diagnosis of T. cati infections and OT. The incorporation of multiple recombinant proteins into multiplexed particle-based flow cytometric assays (e.g., Luminex®) or multiplexed lateral-flow assays would inform antigen selection for optimal sensitivity and specificity. Measurement of specific Ig subclasses or isotypes may also improve the assay performance of recombinant protein-based assays. More studies are needed that focus on the use of antigen detection or avidity for differentiation of current and resolved infections. Well-characterized sera must be collected and shared for meaningful assay development, and reactivity in the TES-Ag IB should be part of the specimen pedigree.

In addition to the development of improved immunodiagnostic methods, other laboratory studies could inform our understanding of the biology of Toxocara and the pathophysiology of the disease. Next-generation sequencing methods should be used to compare the genomes of T. canis larvae recovered from cases of VT and OT. The fact that the clinical signs and symptoms of VT are usually absent in OT (e.g., eosinophilia, elevated serum IgE) suggests that the two syndromes may be caused by different organisms or genotypes. If there is no genetic distinction, a comparison of the proteomes of the larvae from VT and OT may suggest the basis for these different clinical syndromes. Molecular or antibody tools should be developed to differentiate T. canis and T. cati in order to estimate the percentage of toxocariasis caused by T. cati. A thorough characterization of the T. canis glycome through modern glycomics protocols may enable the identification of novel carbohydrate structures that could be targets for new antigen detection methods or differentiation of Toxocara species. Comparative genomics applications could be used to investigate the diversity of Toxocara species from various geographic areas and hosts. The degree of inter- and intraspecies polymorphism in T. canis and T. cati is unknown. Genomic data would appear to be especially valuable to the veterinary pharmaceutical community, which has the resources required to sequence, assemble, and annotate the nuclear genomes of T. canis and T. cati.

Much work remains to be done. Toxocariasis is increasingly recognized as an important zoonosis [69], and as such, improved laboratory tools will be the basis for better understanding of the parasite and the diseases it causes. Comparative genomics, proteomics, and glycomics will result in the identification of novel diagnostic targets and methods for molecular epidemiological studies. The use of improved immunodiagnostic tools will not only enable greater diagnostic accuracy, but will also generate reliable seroprevalence data. Ultimately, these efforts will lead to better strategies to control this preventable disease.