| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Previous Article | Next Article ![]()
Journal of Clinical Microbiology, February 2007, p. 438-442, Vol. 45, No. 2
0095-1137/07/$08.00+0 doi:10.1128/JCM.01735-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
Section of Parasitology, Department of Medical Microbiology,1 Division of Infectious Diseases, Tropical Medicine and AIDS, Department of Internal Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands,3 Harbor Hospital and Institute of Tropical Diseases, Rotterdam, The Netherlands,2 Laboratory of Parasitology and Department of Infectious Diseases, Saint Louis Hospital, Assistance Publique-Hôpitaux de Paris, Paris, France,4 Prince Leopold Institute of Tropical Medicine, Central Laboratory of Clinical Biology, Antwerp, Belgium5
Received 22 August 2006/ Returned for modification 10 October 2006/ Accepted 21 November 2006
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Serological methods to detect Strongyloides infection have been studied previously (2, 6, 10, 11, 20, 23, 24, 26). The most convenient and widely used method is the enzyme-linked immunosorbent assay (ELISA), which detects serum immunoglobulin G (IgG) against a crude extract of infective larvae. However, the ELISA is somewhat cumbersome and labor-intensive, and the test requires a certain level of laboratory infrastructure for performance and interpretation of results, factors that have hampered its applicability in areas where Strongyloides is endemic (29). In contrast to ELISA, dipstick assays are fast and easy to perform (13, 25). Studies with other infections demonstrated a high degree of concordance between the results of ELISA and those of the dipstick assay (4, 13, 25).
In general, serological assays are easier to perform and require more easily obtainable patient material (serum versus fresh stools) than parasitological assays. However, most assays are homemade and require advanced laboratory infrastructure for adequate performance. Therefore, the availability of commercially available serological tests would be advantageous.
In this study, we evaluated a homemade ELISA (Academic Medical Center ELISA [AMC-ELISA]) and an easy-to-perform dipstick assay, based on S. stercoralis antigens, for the serodiagnosis of strongyloidiasis. In addition, two commercial ELISAs (IVD-ELISA and Bordier-ELISA), which were recently introduced, were evaluated.
(Part of this work was presented as a poster at the Annual Spring Meeting of the Dutch Society for Medical Microbiology, April 2006, Papendal, The Netherlands.)
| MATERIALS AND METHODS |
|---|
|
|
|---|
A total of 99 sera from patients with Strongyloides infection were used: 90 from patients suffering from intestinal S. stercoralis infection, which was parasitologically proven by the Baermann method, and 9 from patients with a clinical diagnosis of larva currens. Sera were collected within 1 month after parasitological diagnosis. Sera were stored at 20°C until use.
A total of 220 sera from patients with or without an infectious disease were used as negative controls. (Table 1). Thirty-seven sera from patients suffering from parasitologically proven filariasis were also used (see Table 3).
|
|
Preparation of dipsticks. The dipsticks were prepared as described previously (13). Two lines were made on a nitrocellulose membrane strip (Schleicher & Schüll Bioscience, Dassel, Germany): one with cultured S. stercoralis antigen and a positive control consisting of human IgG (Nordic Immunological Laboratories, Tilburg, The Netherlands). After incubating the membrane strips for 1 h at room temperature, the membrane strips were blocked (to prevent nonspecific binding of protein) by incubating the strips for 30 min in PBS containing 3% (wt/vol) chicken egg white (PBS-3% egg) (Sigma-Aldrich, Inc., St. Louis, MO). After blocking, the membrane strips were washed three times with PBS for 5 min. The dipsticks were stored under dry conditions at room temperature.
Use of dipstick. The dipsticks were incubated for 15 min with serum. After washing three times with TNT buffer (20 mM Tris, 0.5 M NaCl, and 0.2% Tween, pH 7.5), the dipsticks were incubated with horseradish peroxidase-conjugated goat anti-human IgG (Nordic) at a 1:200 dilution in TNT-0.5% egg and subsequently washed three times and placed in a substrate solution containing 3,3-diaminobenzidine tetrahydrochlorid and H2O2 (ICN Biomedical, Inc., Aurora, OH) for 2 min. Rinsing with tap water stopped the reaction. A brown line at the position of the positive control indicated successful processing, while a brown line at the position of the antigen indicated the presence of S. stercoralis-specific antibodies in the tested serum.
Specific-antibody ELISA (AMC-ELISA). Flat-bottomed polystyrene high-bind microplates with small wells (Corning, Inc., Acton, MA) were used. Wells were coated overnight at 4°C with S. stercoralis antigen in 50 mM carbonate buffer (pH 9.6). Subsequently, the wells were treated with PBS-1% egg at 37°C for 30 min. After washing three times with PBS-0.05% Tween 20, serum samples were added for 1 h at 37°C. After washing five times, bound antibodies were incubated with a 1:1,000 dilution of horseradish peroxidase-conjugated goat anti-human IgG (Nordic) in PBS-1% egg for 30 min at 37°C. Subsequently, the wells were washed five times, and substrate solution (0.1% 5-amino salicylic acid in phosphate buffer, pH 5.95, with 0.03% H2O2; Merck, Whitehouse Station, NJ) was added for 1 h in the dark at room temperature. The reaction volume of each reagent at each step was 30 µl per well. The optical density (OD) was read at 492 nm on a Multiskan ascent reader (Labsystems, Helsinki, Finland). A cutoff value was calculated from the mean OD plus twice the standard deviation of the 220 control sera.
Setup of AMC-ELISA and dipstick assay. For the AMC-ELISA and dipstick assay, the optimal concentration of S. stercoralis antigen was determined by making serial dilutions: concentrations of 2 µg/ml and 100 µg/ml gave the best results, respectively. The optimal dilution of test serum for both assays was 1:200 and 1:50, respectively. Testing was done with sera from patients with parasitologically proven S. stercoralis infection and from healthy blood donors.
Commercial ELISAs. Two commercially available ELISAs were studied. In the Strongyloides Serology Microwell ELISA kit (IVD Research, Inc., Carlsbad, CA), human IgG antibodies react with S. stercoralis antigen, and in the ELISA from Bordier (Strongyloides ratti; Bordier Affinity Products SA, Crissier, Switzerland), specific IgG reacts with Strongyloides ratti somatic larval antigens. Both tests were used according to the instructions of the manufacturers. Briefly, for the IVD-ELISA, serum samples were diluted 1:64 in dilution buffer and incubated for 10 min in Strongyloides antigen-coated wells. After washing three times with wash buffer, 2 drops of peroxidase-conjugated protein A were added for 5 min. Subsequently, the wells were washed again three times, 2 drops of chromogen (tetramethylbenzidine) were added for 5 min, and the reaction was stopped with 2 drops of stop solution (1 M H3PO4). The OD was measured at 450 nm with 620 to 650 nm as reference wavelengths. An OD of >0.200 indicated a positive result. For the Bordier-ELISA, serum samples were diluted 1:200 in Tris-buffered saline-Tween solution and incubated for 30 min in S. ratti somatic larval antigen-coated wells. After four washes, bound activity was detected with protein A conjugated to alkaline phosphatase. Four washes, incubation with phosphatase substrate, and stopping with K3PO4 completed the reaction. Absorbance was measured at 405 nm. An OD higher than that of the weakly positive control was considered a positive result.
| RESULTS |
|---|
|
|
|---|
The sensitivity was calculated by testing 99 sera from patients with S. stercoralis disease: 90 patients with intestinal infection and 9 patients with larva currens. Sensitivities of the AMC-ELISA, the dipstick assay, the IVD-ELISA, and the Bordier-ELISA for intestinal infection were 93% (84/90), 91% (82/90), 89% (80/90), and 83% (75/90), respectively (Table 2). In all tests, eight sera from nine larva currens patients gave positive results.
|
|
| DISCUSSION |
|---|
|
|
|---|
In the homemade AMC-ELISA and dipstick assay, S. stercoralis antigen was used because several dominant immunogenic epitopes are not present in Strongyloides cebus, S. ratti, and Strongyloides venezuelensis (7, 10, 22). However, large amounts of S. stercoralis antigen are difficult to obtain: stools of infected patients are needed for collection and culture. We therefore developed an economic ELISA method with microtiter plates with smaller wells (i.e., 30 µl as opposed to 300 µl), resulting in a 10-fold reduction in antigen use (0.12 µg versus 1.2 µg). Isolation and culture of S. stercoralis from stools from one patient resulted in amounts of antigen that were sufficient to perform more than 100,000 ELISAs. The high sensitivity of the AMC-ELISA for intestinal strongyloidiasis (93%) agreed with the results of previous studies carried out with standard microtiter plates in areas where strongyloidiasis is endemic and in areas where it is not endemic (2, 6, 10, 11, 20, 23, 24, 26).
Both high and low specificities with ELISA have previously been reported (2, 6, 8, 10, 11, 20, 23, 24, 26). Because of these variable results, we included a large number of control sera from patients suffering from different parasitological, bacterial, and viral infections; from patients with high titers of antinuclear antibodies and rheumatoid factor in their blood (these sera are known to cause aspecific reactions in serological tests); and from healthy blood donors to determine the specificity. Despite this variety in control groups, a high specificity was achieved. Low cross-reactivity was observed among sera from patients with echinococcosis (AMC-ELISA and dipstick assay) and schistosomiasis (all assays). These false-positive reactions could also be caused by occult strongyloidiasis.
To our knowledge, the dipstick technique reported in this study has never been used for the detection of antibodies against S. stercoralis. For serodiagnosis of other parasitic infections, this dipstick technique has proven to be highly sensitive and specific (4, 13, 14, 30, 31). The dipstick assay proved to be easy to fabricate, easy to use, and rapid (results in less than 1 h). The absence or presence of a line at the position where the antigen was incubated was unambiguous. Both sensitivity and specificity were high (91% and 97.8%, respectively), and the agreement with the AMC-ELISA was high (
= 0.927). The advantage of a dipstick assay is that it is easier to perform for less trained technicians and less equipped laboratories. Drawbacks are the use of more antigen compared to the AMC-ELISA (1 µg versus 0.12 µg) and the inability to quantify test results. For this reason, it is suggested to use the dipstick assay for fast diagnosis of strongyloidiasis and to use the AMC-ELISA for follow-up after treatment.
Eight of nine larva currens cases were positive by both the AMC-ELISA and the dipstick assay. All these cases were negative after three consecutive microscopic stool examinations by the Baermann method, suggesting that a single serological examination is more sensitive for the diagnosis of larva currens than the Baermann method, which is considered one of the most sensitive methods of stool examination (9, 27, 28).
The sensitivity of the Strongyloides Serology Microwell ELISA kit from IVD research, Inc. (IVD-ELISA), was slightly lower, and the specificity was in between those for the AMC-ELISA and dipstick assay (Table 2). The agreement between both ELISAs was very good (
= 0.875). The IVD-ELISA was very rapid (less than 1 h) and easy to perform. For the production of this kit, the soluble fraction of S. stercoralis L3 filariform larvae is used (IVD, personal communication). However, further details about the source, processing, and purification of antigens were not available. In addition, the package inserts of this kit are inconsistent: different cutoff values for the ELISA are recommended for different batches of the kit. For the evaluation in this study, we used kits from one batch with a cutoff value of 0.200. A kit from a later batch with a cutoff value of 0.400 was also evaluated: when tested, a slightly higher specificity was found, but the sensitivity was reduced dramatically (Table 2). The IVD-ELISA has not been cleared by the FDA.
In the Bordier-ELISA (Bordier Affinity Products SA), Strongyloides ratti is used as an antigen. The test has a lower sensitivity than the other assays but had an excellent specificity. The test is also easy to perform but takes longer than the IVD-ELISA (2 h). The agreement between the AMC-ELISA and the Bordier-ELISA and between both commercial ELISAs was very good (
= 0.851 and 0.870, respectively). The use of somatic Strongyloides ratti larval antigens could explain the slightly lower sensitivity of this test because of the incomplete cross-reactivity of both Strongyloides species. This kit has also not been cleared by the FDA.
Cross-reactivity between filarial and Strongyloides antigens is a well-known phenomenon that has been described previously (6, 7, 17-19). With the four tests evaluated in this study, we also observed frequent positive reactions among sera from patients with filarial infections (Table 3). This strongly suggests cross-reactivity, although double infections cannot be excluded. For calculations of the cutoff and specificity of the assays, we excluded results from sera of filarial patients. The antigens of the IVD-ELISA appeared to be less cross-reactive than antigens used in the other assays, especially for Wuchereria bancrofti antibodies. It has been reported that problems of cross-reactivity can partially be overcome by the preincubation of sera with Onchocerca gutturosa extract to deplete the sera from antifilarial antibodies (6, 15) or by using recombinant S. stercoralis antigens for coating of the microtiter plates (21).
When patients have positive Strongyloides serology and originate from or have a history of travel to areas in which filariasis is endemic, it is recommended to run a test for filaria antibodies simultaneously. In our experience, patients with strongyloidiasis have higher antibody levels with Strongyloides antigen than with filarial antigen in ELISAs, an observation that was reversed in patients with filariasis (data not shown).
In the design of our study, parasitologically proven patients were chosen as the "gold standard." Although in this design, it is not possible to prove the superiority of the evaluated tests, two observations from the Harbor Hospital support this suggestion. First, in a group of 40 patients with positive serology, parasitological confirmation was obtained with one, two, or three Baermann examinations in 30, 7, and 3 patients, respectively. Second, among 130 patients who were clinically strongly suspected to have intestinal strongyloidiasis, Baermann examinations were positive in 8 (6%) cases, while serology was positive in 23 (18%). False positivity as an explanation for this is unlikely, since the number of positives is much higher than that in our control group. False positivity due to filarial cross-reactivity is also unlikely, since filariasis is most uncommon in our patient population. These findings suggest that serology might be more sensitive than Baermann examinations. In addition, serology is cheaper and less cumbersome for both the patient and the laboratory.
In summary, the four assays tested are all highly sensitive and specific and can easily be implemented in routine laboratory diagnostics. Tests with S. stercoralis antigen in general have better sensitivity. Cross-reactivity of antifilarial antibodies should be considered for patients from areas of endemicity. Both the IVD-ELISA and Bordier-ELISA show promising results in this first-time evaluation. However, further evaluations are needed, especially with respect to the reproducibility of different batches of the IVD-ELISA.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Published ahead of print on 6 December 2006. ![]()
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Antimicrob. Agents Chemother. | Clin. Microbiol. Rev. |
|---|---|
| Clin. Vaccine Immunol. | ALL ASM JOURNALS |
|---|