Subarachnoidal and intraventricular human neurocysticercosis: application of an antigen detection assay for the diagnosis and follow-up

Authors


Corresponding Author Agnès Fleury, Instituto Nacional de Neurología y Neurocirugía, Insurgentes Sur 3877, Col. La Fama, C.P 14269, México, D.F. México. Tel.: +52 55 5606 4040; Fax: +52 55 5622 3369; E-mail: afleury@biomedicas.unam.mx

Summary

Background  Neurocysticercosis (NC) is a parasitic disease of the central nervous system caused by the larval stage of Taenia solium. Although imaging studies are recommended for diagnosis and follow-up of patients, their high cost and restricted availability limit their use. Among various immunological tests, the detection of HP10 antigen in cerebral spinal fluid (CSF) has proved to be a useful tool for the diagnosis of NC in the case of viable but not dead parasites.

Objectives  This study was designed to evaluate the usefulness of the detection of HP10 antigen for the diagnosis and follow-up of NC patients.

Methods  The effectiveness of this HP10 assay was analysed for the CSF of 46 confirmed NC cases (21 men, 25 women) who had been clinically and radiologically described.

Results  In 21 of 24 NC patients (87.5%) with viable parasites localized in the SA space at the base of the brain or in the ventricles these were detected by means of the HP10 assay, whilst none of the three patients with viable parasites in the parenchyma or sulci had these detected. Used for the follow-up of patients after cysticidal treatment, it was showed that levels of HP10 dropped significantly only among those patients whose cysticerci were clearly damaged.

Conclusions  HP10 antigen assay is recommended as a support for diagnosis and follow-up in NC patients with viable parasites localized in the SA space at the base of the brain or in the ventricles, thereby potentially reducing the number of imaging studies required.

Abstract

Données de base  La neurocysticercose est une maladie du système nerveux, causée par Taenia soliumà son stade larvaire. Bien que les méthodes d'imagerie soient recommandées pour le diagnostic et le suivi des patients, leur coût élevé et leur disponibilité restreinte limitent leur application. Parmi divers tests immunologiques, la détection de l'antigène HP10 dans le fluide cérébral spinal s'est révélée un outil utile pour le diagnostic de la neurocysticercose dans le cas de parasites viables mais non-dans celle de parasites morts.

Objectifs  Evaluation de l'utilité de la détection de l'antigène HP10 dans le diagnostic et le suivi des patients avec une neurocysticercose.

Méthodes  L'efficacité du test HP10 a été mesurée sur le fluide cérébral spinal de 46 cas de neurocysticercose (21 hommes et 25 femmes) cliniquement et radiologiquement confirmés.

Résultats  Sur 24 patients avec des parasites viables, les parasites ont été détectés par le test HP10 chez 21 (87,5%) patients chez qui ils étaient localisés dans l'espace SA à la base du cerveau ou dans les ventricules. Chez les 3 autres patients, le test n'a pas détecté de parasites viables localisés au niveau du parenchyme ou des sulci. L'utilisation du test dans le suivi des patients traités avec de cysticides a démontré que le taux de HP10 chutait significativement, seulement chez les patients chez qui les cysticerques avaient été clairement endommagés.

Conclusions  Le test de l'antigène HP10 est recommande comme support au diagnostic et au suivi des patients avec une neurocysticercose à parasites viables localisés dans l'espace SA à la base du cerveau ou dans les ventricules. Cela réduira le nombre de tests d'imagerie nécessaires.

Abstract

Antecedentes  La neurocisticercosis (NC) es una enfermedad parasitaria del sistema nervioso central causada por el estado larval de Taenia solium. Aunque se recomiendan estudios de imagen para el diagnóstico y seguimiento de los pacientes, su alto coste y la disponibilidad restringida limitan su uso. Entre varias pruebas inmunológicas, la detección del antígeno HP10 en el líquido cefaloraquídeo (LCR) ha demostrado ser una herramienta útil para el diagnóstico de la NC en caso de existir parásitos viables, más no en el caso de parásitos muertos.

Objetivos  Este estudio fue diseñado para evaluar la utilidad de la detección del antígeno HP10 en el diagnóstico y seguimiento de pacientes con NC.

Métodos  La efectividad del ensayo con HP10 fue analizado en el LCR de 46 casos confirmados de NC (21 hombres, 25 mujeres), con diagnóstico clínico y radiológico.

Resultados  A 21 de los 24 pacientes con NC (87.5%) con parásitos viables localizados en el espacio subaracnoidal (SA) en la base del cerebro o en los ventrículos, se les habían detectado por medio del ensayo de HP10, mientras que a ninguno de los tres pacientes con parásitos viables en el parénquima o en los surcos (sulci) les habían sido detectados. Utilizado para el seguimiento de pacientes después del tratamiento cisticida, se demostró que los niveles de HP10 caían significativamente solo entre aquellos pacientes en los que el cisticerco estaba claramente dañado.

Conclusiones  El ensayo de antígeno HP10 está recomendado como un apoyo en el diagnóstico y seguimiento de pacientes con NC con parásitos viables localizados en el espacio SA en la base del cerebro o en los ventrículos, reduciendo potencialmente el número de estudios por imagen requeridos.

Introduction

Human cysticercosis, highly prevalent disease in many developing countries, is frequently located in the central nervous system. The latter, known as neurocysticercosis (NC), manifests a large variety of symptoms with levels of severity ranging from subclinical forms, headaches and seizures, to severe intracranial hypertension (Del Brutto et al. 1988; Sciutto et al. 2000; Fleury et al. 2004; Wallin & Kurtzke 2004; Del Brutto 2005). This heterogeneity hinders a reliable diagnosis when this is based only on the clinical picture.

Thus, diagnosis still requires image studies [computed tomography (CT scan) and magnetic resonance image (MRI)] as a back up in order to demonstrate the presence and location of parasites and the intensity of the associated inflammatory reaction. These studies provide the necessary information for selecting the most appropriate treatment and follow-up schedule. This parasitosis is also characterized by its heterogeneous evolution: some cases will resolve without the need of cysticidal drugs while others require them. In this group of patients, some need only one course of treatment, while others require several treatment schedules. Consequently, the rule in modern hospitals is to proceed with further confirmatory imaging studies following specific treatment.

The high cost and inaccessibility of imaging studies in rural areas has stimulated the development of immunodiagnostic tests for NC by means of the detection of antibodies or antigens in serum and/or cerebrospinal fluid (CSF) (Rosas et al. 1986; Tsang et al. 1989; Larralde et al. 1990; Ramos-Kuri et al. 1992; Garcia et al. 1995, 2000; Verastegui et al. 2003).

Unfortunately, antibodies persist for extended periods in serum and CSF, long after the parasite has been destroyed (Fleury et al. 2003a) and so their detection is not helpful for designing follow-up. Contrastingly, preliminary results indicate that the detection of an antigen secreted by viable parasites (Harrison et al. 1989) may be useful for diagnosis and follow-up of NC after drug treatment, thereby reducing the number of imaging studies required (Garcia et al. 1998, 2000, 2002; Ferrer et al. 2002; Fleury et al. 2003b; Nguekam et al. 2003). These studies did not however explore the relevance of the variable locations of parasites in the brain, as affecting the sensitivity of the HP10 assay. This could be an important consideration as variations in the location of cysticerci in the CNS represent a critical factor in terms of the clinical and immunological heterogeneity of the disease (Fleury et al. 2004).

This study was therefore designed to evaluate the effectiveness of HP10 antigen detection for NC diagnosis in patients with parasites at different stages of degeneration and in varying location. Its performance in the follow up of patients after cysticidal treatment was also considered.

Materials and methods

Cerebrospinal fluid samples and definition of cases

Sixty CSF were obtained by means of lumbar puncture from 46 NC cases (21 men and 25 women) at the National Institute of Neurology and Neurosurgery in Mexico City (INNN). Age at diagnosis ranged from 13 to 67 years. Samples of CSF were obtained from 14 patients before and after cysticidal treatment, from 18 others only after cysticidal treatment and from yet another 14 only before cysticidal treatment. Diagnosis was made based on: clinical manifestations (presence of seizures, focal deficit and intracranial hypertension) and imaging studies (image compatible with the presence of vesicular, colloidal or calcified cysticerci). Patients included in this study had symptomatic NC defined by the presence of one or more of the clinical manifestations mentioned above and presented images compatible with the presence of cysticerci in the CNS as defined by two neurologists from the INNN. The cysticidal drug used was albendazole at doses of 15–30 mg/kg/day over 10 days. It was administered with corticosteroids (from 20 to 50 mg of prednisone per day for between 10 days up to 2 months) in order to reduce the inflammatory reaction. The treatment regime varied from one case to another, mainly depending on the intensity of the inflammatory reaction. Treatment success was evaluated 3 months later using a CT scan and/or MRI.

The following information was collected for each NC case from the radiological studies (CT scan and MRI): number of lesions (single vs. multiple), state of the cysticerci (vesicular, colloidal or calcified) and CNS location. CNS location comprised those that were located at the subarachnoid space at the base of the brain (SA base) or at the cortical sulci (SA sulci), parenchyma, or ventricle (IV).

The CSF cellularity was recorded (cells were counted in a Neubauer chamber, most of them were lymphocytes and eosinophils were detected in only a few cases). Cellularity was considered to have increased when the number of cells exceeded 5/ml. This study was approved by the Local Hospital Review Board (LHRB). Patients were informed that the CSF samples obtained during their hospital studies would be used for this study and gave their consent.

This study had a double blind design. Neither neurologists nor radiologists knew the results of the HP10 test and laboratory technicians did not know the diagnosis of the neurological patients.

Detection of parasite antigens

Parasite antigens were detected by means of Ag-ELISA as described previously (Garcia et al. 1998, 2000; Fleury et al. 2003b). ELISA plates were routinely set up to include controls, two standard positive samples (NC confirmed cases) and nine non-NC CSF from neurological patients in order to correct for small day to day variations. Samples were run in duplicate and repeated at least twice.

Briefly, Immulon I plates (Nalge, Nunc International, Rochester, New York, USA) were coated with mAb HP10 (100 μl at 10 μg/ml in 0.07 m NaCl buffered with 0.1 m borate, pH 8.2) and left overnight at 4 °C. The plates were washed three times for 5 min with 200 μl/well of wash solution (0.9% w/v NaCl containing 0.05% v/v Tween 20). The plates were blocked using 200 μl diluent (phosphate-buffered saline containing bovine serum albumin 1.0% w/v and 0.05% v/v Tween 20) and left for 60 min at room temperature before being washed in a similar way. Undiluted CSF samples (100 μl/well) were added and incubated for 30 min at 37 °C. Bound HP10 parasite antigens were detected using biotinylated mAb HP10 (2.5 μg/ml in diluent, for 30 min at 37 °C), horseradish peroxidase-conjugated streptavidin (Amersham Ltd, Piscataway, New Jersey, USA; 1:2000 in diluent, 30 min at 37 °C) and tetramethylbenzidine (Kirkegaard Perry Laboratories, Gaithersburg, MD, USA) as substrate. The colour reaction was allowed to proceed for 5 min at 37 °C in the dark and was halted by adding 100 μl 0.2 m H2SO4 (Baker). Optical density (OD; 450 nm) was determined in an ELISA processor (Human GmbH, Wiesbaden, Germany, Humareader, Model 2106).

A sample was considered positive if the mean OD value was greater than the cut-off value, which was defined as 0.34 (corresponding to the mean of the OD of 9 negative CSF samples + 2 SD. Between plates, no-significant changes in the cut-off values were obtained which ranged from 0.32 to 0.34. It should be noted that five of the control samples were inflammatory having a CSF cellularity higher than 5 cells/ml (2 from a case with viral encephalitis and the other three from a case with silvius aqueduct stenosis of infectious origin), which may be related to the high cut-off value.

Statistical analysis

Data were processed in Excel 7.0 (Microsoft) and SPSS 10.0 for Windows. Nonparametric descriptions [median and interquartile range (IQR)] were calculated. The Mann–Whitney U nonparametric test and univariate analysis of the data (Pearson's X2 test with Yates's correction or Fisher test when appropriate) were used to identify the differences in absorbance values and CSF cellularity between groups. P ≤ 0.05 was considered significant. Multivariate analysis was conducted using forward stepwise logistic regression analysis. Only variables reaching P ≤ 0.4 in the univariate analysis were considered for the multivariate analysis. Odds ratios (OR) and confidence intervals were determined.

Results

Of the sixty CSF samples, imaging studies allowed unequivocal determination of the viability status of cysticerci in only 51 samples, even with the use of MRI, which permits vesicular cysticerci to be distinguished more precisely.

HP10 antigen level and parasite viability

The data presented in Figure 1 indicates HP10 levels among patients with vesicular cysts, detectable by means of image analysis either before or after cysticidal treatment and among post-treatment patients with either damaged cysticerci or without visible cysticerci. From the 14 cases in which pre- and post-treatment samples were available, only pre-treatment samples were included. Patients with damaged cysticerci presented a significant reduction in HP10 antigen (median = 0.29; IQR 0.23–0.33) when compared with patients with vesicular parasites (median = 0.56; IQR 0.4–1), P < 0.0001. Few NC vesicular patients presented low levels of HP10 leading to an overlapping of the quartile ranges (Figure 1).

Figure 1.

 Individual HP10 levels (OD) in 39 CSF from the NC patients with vesicular or damaged (colloidal, calcified or not detected) cysticerci. Only cases in which viability of cysticerci was unequivocally determined are included and only one CSF by patient was included. The bar represents the mean of each group. Open circles represent CSF from patients with vesicular cysts who did not receive any treatment. Black circles represent HP10 levels after treatment of vesicular cysticerci. Black triangles represent CSF with damaged cysts after treatment.

HP10 level of pre-treatment vesicular cases (median = 0.55; IQR 0.4–1) did not differ significantly from the post-treatment vesicular cases (median = 0.62; IQR 0.23–0.96), P = 0.68.

HP10 level and parasite location

To further explore the factors involved in the heterogeneous levels of HP10 detected in vesicular NC patients (Figure 1), parasites were localized using image analysis (summarized in Table 1). As shown here, the level of HP10 in CSF samples from the 46 patients included in this study strongly depended on the location of parasites in the CNS. Higher levels of HP10 were detected in the CSF of patients with cysts in the subarachnoidal space at the base of the brain (SA base) or in the ventricles (IV) than in parenchyma (P < 0.0001 and P = 0.12, respectively; Table 1). The sensitivity of this assay for detection of vesicular parasites located in SA space at the base of the brain or in the ventricles was of 87.5% (21 of 24), whilst none of the three vesicular parasites located in parenchyma or cerebral sulci were detected. In spite of the low number of samples, the differences in the sensitivity of the two compartments were significant (P = 0.007). It is important to note that most of the patients with SA base and IV viable cysticerci exhibited an increase in cellularity in the CSF. However, we did not find any correlation between HP10 levels and the number of cells presented in the CSF (r = −0.06, P = 0.77).

Table 1.   Level of HP10 detected according to the NC status
 No. of patientsCSF cellularity (cells/ml) [median (IQR)]HP10 [median OD (IQR)]
  1. SA, subarachnoidal; IQR, interquartile range.

SA base2116 (9.5–48)0.62 (0.42–1.2)
 Vesicular parasite1918 (10–52)0.65 (0.48–1.3)
 Damaged only25.5 (–)0.34 (–)
Ventricular727 (5–144)0.43 (0.28–0.7)
 Vesicular parasite533 (27–410)0.54 (0.3–0.85)
 Damaged only22.5 (–)0.29 (–)
SA sulci only or parenchyma only81.5 (0–7.5)0.29 (0.16–0.33)
 Vesicular parasite32 (–)0.28 (–)
 Damaged only50.001 (0–19.5)0.29 (0.19–0.32)
Without distinguishable cysticerci (post-treatment)1022 (0–103)0.29 (0.19–0.36)

The usefulness of HP10 in the follow-up of neurocysticercosis patients

Of the 14 patients with pre- and post-treatment samples, 10 pre-treatment patients were positive in terms of identification by HP10. Among all these paired patients, parasites were located in the SA space at the base of the brain or in the ventricles. As shown in Table 2 and in Figure 2, effective cysticidal treatment of NC patients with vesicular cysticerci and high levels of HP10 antigen resulted in radiological evidence of cysticercal damage. Besides this a marked decrease of HP10 antigen levels in the CSF as early as 4 months after treatment (Table 2, patient nos 1–6) was observed. In other patients after 1 month of treatment, no major changes in these two parameters were found (Table 2, patient nos 7–9). Only in one patient was HP10 antigen detected after treatment, even when parasites could not be detected by image analysis (Table 2, patient no. 10).

Table 2.   Levels of HP10 antigen in CSF of NC patients with cysts in the subarachnoidal space at the base of the skull or in the ventricles before and after cysticidal treatment
Patient no.Before (IS)After (IS, months of treatment)
  1. * HP10 level (OD).

  2. IS, state of the cysticerci by imaging studies (V, vesicular; Ca calcified; Co colloidal; ND not detected).

 10.65* (V)0.11 (ND, 23)
 20.69 (V)0.20 (ND, 9)
 30.43 (V)0.16 (Ca, 4)
 41.00 (V)0.49 (Ca, 12)
 50.54 (V)0.37 (Co, 4)
 61.28 (V)0.28(Ca, 13)
 71.05 (V)0.90 (V, 1)
 80.88 (V)0.50 (V, 1)
 91.34 (V)1.11 (V, 1)
100.56 (V)0.49 (ND, 7)
Figure 2.

 MRI and HP10 determinations of two NC cases before (a and c) and after (b and d) cysticidal treatment. (a) and (b) correspond to cases with intra-ventricular cysticerci (case no. 5 in Table 1), while (c) and (d) correspond to cases with parasites located in the subarachnoidal space (SA) at the base of the brain (case no. 6 in Table 1). Arrowheads indicate places in which lesions or cysts were found.

HP10 level and gender

Following previous findings which indicated that women NC patients exhibit more inflammation as indicated in imaging studies (Trelles 1992) as well as higher cellularity (Fleury et al. 2004), HP10 antigen levels were compared between the two genders. However, no differences in the median of HP10 levels between women and men were found [0.34 (IQR 0.29–0.98) vs. 0.40 (IQR 0.22–0.52)], P = 0.28. The sensitivity of the assay was shown to be higher among women than among men (92.3%vs. 64%) albeit at the limit of significance (P = 0.07).

Neither did we find any relation between age and results of the HP10 test: in HP10 positive subjects, median age was 38 years (IQR 33–50), while in HP10 negative subjects, median age was 35 (IQR 33–49.5), P = 0.18.

Variables relating to HP10 positive levels

To determine which factor(s) might be associated with the presence of the HP10 antigen, a multivariate analysis was conducted. HP10 levels were considered to represent the dependent variable; and age, gender, CSF inflammation, parasite localization and viability to represent the independent variables. In this analysis, the relevance of cyst localization in the SA base of the brain [OR = 7.4 (0.9–59) P = 0.06], inflammatory CSF [OR: 30 (3.1–285), P = 0.003] and presence of viable parasites [OR = 37 (2.7–519), P = 0.007] were found to be strongly associated with positive HP10 levels.

Specificity of the HP10 assay for detection of active neurocysticercosis

The specificity of the HP10 assay for identifying NC cases with viable parasites was 96.1% (25 of 26). Only one case without any viable cyst being detectable by means of MRI, proved positive in the HP10 antigen assay. This case received specific treatment 1 year before the lumbar puncture, and yielded inflammatory CSF.

Discussion

In this study the usefulness of the HP10 antigen detection assay in CSF for the immunodiagnosis and follow-up of NC cases was confirmed for a cohort of clinically and radiological defined neurocysticercotic patients. As we reported previously, increased levels of the cysticercal secreted HP10 antigen were detected when vesicular parasites were identified in the CNS by means of neuroradiological studies (CT scan or/and MRI; Fleury et al. 2003b). However, a wide-range of antigen levels ranging from OD of 0.15 to 1.6 was detected among NC patients. In this study, we contribute statistical evidence, which indicates that localization of parasites and their state, as well as the level of CSF inflammation are the most important variables relating to HP10 levels.

The highest levels of HP10 antigen in CSF were obtained when viable parasites were located in the SA space at the base of the brain or in the ventricles. This is an important finding as parasite location in these sites produces the most severe forms of the disease and is of high frequency in Mexico and Latin America (Fleury et al. 2004; Del Brutto 2005). Regarding the relationship between CSF inflammation and HP10 levels it is interesting to note that CSF inflammation may be promoted by the high level of antigen secreted in the CSF when the parasite is located in the SA space at the base of the brain or in the ventricles. Thus, HP10 may enhance IL5 and IL6 interleukins, which are found to contribute to the local inflammation, which frequently accompanies these severe cases of NC (Chavarria et al. 2005). Disregarding the causes underlying the CSF inflammation, based on the information shown here, it is feasible to propose that the level of HP10 antigen in the CSF is an additional criteria for diagnosing patients with vesicular forms of NC. This might be particularly useful in cases where the parasite is located in the SA space at the base of the brain or in the ventricles, because of the fact that radiological studies (particularly CT scan) do not always provide a certain diagnosis (Llompart Pou et al. 2005).

With respect to the usefulness of HP10 for the follow-up of the NC patients it is worth noting that levels of HP10 dropped significantly only among those patients where the cysticerci was clearly damaged as indicated by radiological studies and shown in Table 2. Contrastingly, levels of remaining HP10 increased among those patients in which vesicular parasites persisted after treatment. Only in one patient did the level of HP10 persist even when cysticerci were no longer detected in the radiological studies. It might be the case that viable cysts remained, but that these were not detectable in the radiological studies.

On the other hand, although cost estimates for the NC patients have not been determined with precision, the cost of imaging studies for the purpose of diagnosis and follow-up treatment for these patients represents an important component. Moreover, these facilities are restricted to specialized hospitals in large urban centres. Thus, many efforts have been made to develop sensitive and specific immunological assays to support the diagnosis and follow-up of the patients (Rosas et al. 1986; Tsang et al. 1989; Larralde et al. 1990; Ramos-Kuri et al. 1992; Garcia et al. 1995, 2000; Verastegui et al. 2003). In this respect, methods based on the detection of antibodies are not useful for defining follow-up because once patients are infected, antibodies can remain in the CSF of the NC patients for months even when the parasite has become calcified (Fleury et al. 2003a). In contrast, we demonstrate in this study that the level of HP10 antigen is highly sensitive to the destruction of the parasite.

Taking these results into consideration, it is now possible to use the CSF HP10 assay as a potent tool for supporting diagnosis and follow-up among the most critical NC patients in neurological institutions. We are now making the necessary arrangements in order to make this assay accessible to consumers in the near future. Results of this assay must now be evaluated in sera, as its application in this area might be of great help in rural conditions where lumbar puncture is not easy to perform.

Conclusions

The HP10 assay is reliable for NC diagnosis and for supporting NC patient follow-up, when parasites are localized in the subarachnoidal space at the base of the brain or in the ventricles, but not when localized in the parenchyma or in the subarachnoidal space of the sulci.

Acknowledgements

We thank Mercedes Baca and Elvira Salazar for technical assistance. Caroline Karslake corrected the English version of the manuscript. This work was supported by Howard Hughes Medical Institute (55000643) and Silanes Laboratory.

Ancillary