Parasite‐Specific IgG Response and Peripheral Blood Eosinophil Count Following Albendazole Treatment for Presumed Chronic Strongyloidiasis

Harin Karunajeewa FRACP, Heath Kelly FAFPHN, David Leslie FRCPA, Jenny Leydon BSc, Pao Saykao MD, Beverley‐Ann Biggs FRACP
DOI: 84-91 First published online: 1 March 2006


Background In developed countries, asymptomatic chronic Strongyloides stercoralis infection occurs in immigrants from endemic regions of the world. Accurate and reliable means of diagnosis and follow‐up are required for effective management. The role of S stercoralis enzyme‐linked immunosorbent assay (ELISA) in this context was examined.

Methods In this study, 95 asymptomatic Laotian immigrants living in Melbourne, Australia, for an average of 12 years, were screened for S stercoralis infection using stool microscopy, eosinophil count, and serology by ELISA. Twenty‐two patients with a positive ELISA were treated with albendazole, 400 mg twice daily for 3 days, and monitored with serology, fecal microscopy, and eosinophil counts at 2, 6, 12, and 36 months after treatment.

Results Patients with moderately reactive baseline ELISA and no eosinophilia had no significant change in either measure over the 36 months of follow‐up. All patients with a strongly reactive baseline ELISA showed a reduction in reactivity over the first 6 months of treatment. However, in 50% of these patients, reactivity increased between 12 and 36 months, suggesting treatment failure and relapse of infection. One patient had confirmed treatment failure based on the development of hyperinfection syndrome.

Conclusion The results support evidence that serology is a valuable tool in monitoring treatment responses in patients with suspected strongyloidiasis and highlights the need to ensure that S stercoralis is completely eradicated after treatment.

Infection with Strongyloides stercoralis is a significant global public health problem, affecting an estimated 30 million people.1,2 Chronic strongyloidiasis is a potential cause of morbidity and mortality in immigrant populations living in developed countries, as well as in other high‐risk groups.1,3–11 Among helminthic infections, S stercoralis is distinguished by its ability to maintain a self‐sustaining life cycle within the human host by production of autoinfective larvae.2,12 Studies in Southeast Asian immigrants and refugees suggest that S stercoralis infection may persist for many years after resettlement, even in those who have had access to screening and treatment programs.4,7,13

Most chronically infected individuals have only mild or no symptoms; however, those who are immunocompromised may develop a hyperinfection syndrome.2,12,14–16 This has high mortality and frequently occurs when the diagnosis of S stercoralis was not suspected prior to the institution of immunosuppressive therapy or when previous treatment for S stercoralis has, in retrospect, been ineffective.12,14,15 A sensitive diagnostic assay and a reliable means of clearly documenting successful treatment is a priority in the management of strongyloidiasis in individuals who may have been exposed many years previously and in whom parasite numbers may be low and larval excretion intermittent.17 Parasitological diagnosis based on fecal examination is insensitive, especially in chronic low‐level S stercoralis infection.2,12,18 Eosinophilia has a good positive predictive value for S stercoralis,13 but its diagnostic value is limited as it is prone to fluctuation and has poor sensitivity.2,7 Consequently, S stercoralis serology is a valuable adjunct in diagnosis, especially in immigrants who have been previously chronically exposed to S stercoralis in an endemic setting.19 However, evaluation of sensitivity and specificity of this test has been limited by the lack of a gold standard comparison and of standardized, commercially available antigen.

We have previously evaluated the prevalence of S stercoralis in Laotian immigrants living in Melbourne, Australia, for an average of 12 years. Patients considered to be chronically infected were identified based on fecal microscopy and/or a positive enzyme‐linked immunosorbent assay (ELISA) together with eosinophilia.13 We hypothesize in this study that, in the absence of positive fecal microscopy, the likelihood of active Strongyloides infection is indicated by the strength of reactivity of the ELISA and an elevated eosinophil count and that such infection may be confirmed by a fall in these values after treatment. Laotian immigrants with positive serology were treated and followed for 3 years to try to confirm those who were infected initially and to monitor the success of treatment.


Patient recruitment

A cross‐sectional survey was conducted in Laotian immigrants attending a general medical practice in Melbourne, Australia, as previously described.13 Briefly, Laotian settlers aged 18 years or older and recorded on the patient register of the general practice were invited to participate. Those who volunteered were screened with a single stool examination, eosinophil count, and S stercoralis ELISA. Those with evidence of presumed infection, defined as a positive ELISA (regardless of whether or not larvae of S stercoralis were visualized on stool examination), were invited to enroll in the follow‐up study.

In addition to those enrolled through the initial survey, additional patients were enrolled through an informal screening process at the same medical practice during the time between the conclusion of the original cross‐sectional survey in June 1998 and the initiation of the current follow‐up study in December 1998. Identical screening methods and criteria for diagnosis of infection were used in these patients. No study participants reported known coinfection with HIV. Ethics approval for the study was obtained from the Melbourne Health Research and Ethics Committee.

Sample collection, treatment, and follow‐up

All laboratory testing was done at the Victorian Infectious Disease Reference Laboratory (VIDRL). Blood samples were collected into plain and ethylenediaminetetraacetic acid tubes for ELISA and eosinophil counts, respectively. In accord with a standard reference range, an eosinophil count of greater than 0.40 × 109/L was considered to be elevated. Diagnostic fecal samples were collected into sodium acetate‐acetic acid–formalin fixative because the volunteers came from a wide geographic area and were not able to return fresh samples. Therefore, Harada–Mori and agar‐plate cultures were not performed. Fecal samples were not collected during follow‐up due to the low number of positive samples in the original study.13 All patients received treatment between December 1998 and February 1999 with albendazole, 400 mg twice daily for 3 days. Albendazole was the standard treatment in Australia for strongyloidiasis during this period as recommended by World Health Organization.20 Patients were asked to reattend 2, 6, 12, and 36 months after initial treatment at which time blood was collected for serology and eosinophil count. At the discretion of the attending physician and independently of this study, five patients also received a single dose of ivermectin (200 μg/kg) between 18 and 26 months after the initial treatment with albendazole.

Strongyloides stercoralis ELISA

Serum was stored at −20°C before being tested for S stercoralis antibodies using an in‐house ELISA based on the method of Genta and Lillibridge,21 with Strongyloides ratti antigen obtained from the State Health Laboratory (Perth, WA, Australia) or the Institute of Clinical Pathology and Medical Research (Sydney, NSW, Australia). This test was originally demonstrated to have a sensitivity and a specificity of 93 and 95%, respectively.22,23 Other nematode infections that may cross‐react in the test, such as filariasis, ascariasis, and acute schistosomiasis, were unlikely as most patients had resided in Australia for many years.

At the time of initiation of the study, VIDRL defined an optical density (OD) of >0.5 as positive for S stercoralis infection, based on a reference range determined from a comparison of the distribution of OD values in uninfected controls with sera of those with microscopically confirmed infection. This cutoff was used to determine inclusion into the study. To eliminate run‐to‐run variation in the ELISA, all sera collected from a patient were tested in parallel on the same run, and results used for analysis were from a single run, with no transformation or averaging.

To verify the consistency of the assay, multiple assay runs were performed over the course of the study period comparing negative, low‐positive, and positive control sera. Performance over a range of serum antibody concentrations was examined using serial dilutions of a positive control serum.

Patient classification

Patients were classified into four groups based on the presence or absence of eosinophilia (>0.4 × 109/L) and/or a strongly positive ELISA OD at baseline. After preliminary analysis, a strong positive ELISA was defined as >1.0. The patient classifications used were eosinophil count normal (≤0.4 × 109/L) and ELISA OD ≤1.0 (group 1), eosinophil count raised and ELISA OD ≤1.0 (group 2), eosinophil count normal and ELISA OD >1.0 (group 3), and eosinophil count raised and ELISA OD >1.0 (group 4).

Data analysis

Geometric mean titers of OD readings and arithmetic mean eosinophil counts were calculated for each of the patient groups. Patient groups were compared on the basis of presumption of infection. Group 1 was presumed to be not infected, and groups 2 to 4 were presumed to be currently infected. For statistical analysis, data from groups 2 to 4 were combined and compared to data from group 1 using the odds ratio (OR) and Fisher's exact test (p < 0.05 was taken as significant) with SPSS software (SPSS Inc., Chicago, IL, USA).


Ninety‐five Laotian settlers participated in the original survey.13 Of these, 22 (23%) had a positive ELISA (OD >0.5) including 2 with larvae of S stercoralis visualized in feces. Of the 22 patients, 6 declined or were unable to participate in the follow‐up study. Subsequent informal screening identified an additional 9 patients with positive serology. Three patients were subsequently lost to follow‐up during the study and were excluded as an insufficient number of blood samples were collected. Of the final study cohort of 22 patients, blood was collected from 19 at 2 months, from 21 at 6 months, from 19 at 12 months, and from 22 at 36 months. One patient died of hyperinfection syndrome 12 months after recruitment and so had only 2‐, 6‐, and 12‐month blood tests.15 . Baseline characteristics and follow‐up blood sample collections for each of the four groups of patients are summarized in Table 1.

View this table:
Table 1

Baseline characteristics and follow‐up blood sample collection for the 22 patients with a diagnosis of Strongyloides stercoralis infection included in this study

Group 1Group 2Group 3Group 4Total
Number of patients742922
Mean age at recruitment, years (range)42.4 (36–57)47.0 (40–56)62.4 (50–74)43.1 (30–58)45 (30–74)
Number of males (%)2 (29)4 (100)1 (50)5 (56)12 (55)
Geometric mean ELISA optical density (± SD)0.67 (0.17)0.65 (0.19)1.31 (0.02)1.80 (0.72)1.07 (0.82)
Mean eosinophil count (± SD)0.17 (0.12)0.57 (0.11)0.26 (0.06)0.7 (0.24)0.37 (0.25)
Positive stool microscopy for S stercoralis01012
  • ELISA = enzyme‐linked immunosorbent assay.

Group 1: OD 0.5 to 1, normal eosinophil count

The seven patients in group 1 had little change in ELISA reactivity or eosinophil count following treatment (Figure 1). Patient 4 had an unexpected rise in eosinophil count to 0.63 at 12 months that was not accompanied by a rise in ELISA reactivity. This had returned to normal when retested at 36 months.

Figure 1

Strongyloides enzyme‐linked immunosorbent assay (ELISA) reactivity (A) and eosinophil counts (B) in patients with an ELISA OD ≤ 1.0 and eosinophil count ≤0.4 (group 1).

Group 2: OD 0.5 to 1.0, raised eosinophil count

All four patients in group 2 had a reduction in antibody levels during the 36 months after treatment (Figure 2). In two of these, reactivity increased immediately after treatment and then gradually declined to the low positive range during follow‐up. One (patient 10) who had a fall in ELISA reactivity immediately following treatment had S stercoralis parasites identified on stool examination at baseline that were not detected thereafter.

Figure 2

Strongyloides enzyme‐linked immunosorbent assay (ELISA) reactivity (A) and eosinophil counts (B) in patients with an ELISA OD ≤ 1.0 and eosinophil count >0.4 (group 2).

Group 3: OD >1.0, normal eosinophil count

Of the two patients in group 3, one had a clear reduction in ELISA reactivity accompanied by an initial fall in eosinophil count (Figure 3). The other had a minor fall in ELISA reactivity over the first 12 months, followed by an increase by 36 months that was accompanied by a rise in eosinophil count from 0 to 0.7.

Figure 3

Strongyloides enzyme‐linked immunosorbent assay (ELISA) reactivity (A) and eosinophil counts (B) in patients with an ELISA OD >1.0 and eosinophil count ≤0.4 (group 3).

Group 4: OD >1.0, raised eosinophil count

All nine patients in group 4 showed a reduction in ELISA reactivity over the first 6 months following treatment (Figure 4A). This was accompanied by a reduction in eosinophil count in all patients except patient 19, in whom eosinophil count initially dropped from 0.75 at baseline to 0.59 at 2 months and then increased to 1.3 at 6 months (Figure 4B). This patient subsequently presented with hyperinfection syndrome after 18 months by which time the ELISA reactivity had risen dramatically, whereas the eosinophil count was normal, probably masked by the inadvertent administration of corticosteroids at 16 months after treatment with albendazole.15 Of the remaining patients in this group, between 6 and 36 months, ELISA reactivity remained stable or dropped in six (patients 15, 16, 18, 20, 21, and 22), rose slightly in one (patient 14), and increased dramatically in one (patient 17). ELISA reactivity remained well above the reference range in all patients up to 36 months after treatment.

Figure 4

Strongyloides enzyme‐linked immunosorbent assay (ELISA) reactivity (A) and eosinophil counts (B) in patients with an ELISA OD >1.0 and eosinophil count >0.4 (group 4).

#Patients 14, 15, 17, and 21 received a single dose of ivermectin between 18 and 26 months after the initial albendazole treatment.

*Patient 19 died of hyperinfection syndrome 18 months after albendazole.

Ivermectin treatment

Four patients in group 4 (patients 14, 15, 17, and 21) and one in group 1 (patient 7) were retreated with ivermectin (200 mg/kg) between 18 and 26 months after initial albendazole treatment at the discretion of the attending physician. Of these five patients, only two had eosinophilia immediately prior to treatment (patients 17 and 21). Eosinophilia resolved in both these cases, but reduction in ELISA reactivity was only minor in one patient (patient 21), whereas the other (patient 17) had a dramatic rise in reactivity following ivermectin, evident at 36 months. Of the remaining three patients treated with ivermectin (all of whom had normal eosinophil counts before treatment with ivermectin), one had a minor fall in ELISA reactivity (patient 7) and the other two had minor rises (patients 14 and 15). Eosinophil counts remained within normal limits in all these patients.

Mean data for groups 1 to 4

Mean values of OD and eosinophil count for each of the four groups are represented in Figure 5. Data from groups 2 to 4 were combined for this statistical analysis. In group 1, there was no fall in the geometric mean OD and only a minor fall in mean eosinophil counts (which remained within normal limits) following treatment, while in groups 2 to 4, both parameters decreased significantly following treatment. For these patients, the major reduction in both ELISA reactivity and eosinophil count occurred in the first 6 months following treatment. Six months after treatment, ELISA reactivity was reduced in three of seven patients (43%) in group 1 compared to 14 of 15 patients (93%) in groups 2 to 4 (OR 18.7; 95% CI 1.12–667.4; p= 0.02, Fisher's exact test). At the same time, eosinophil count had fallen in two of six patients (33%) in group 1 compared to in 13 of 15 patients (87%) in groups 2 to 4 (OR 13.0; 95% CI 0.96–286; p= 0.03). There was no difference in age or any known differences in immunological or nutritional status in those who had a sustained serological response after treatment compared with those who did not.

Figure 5

(A) Geometric mean Strongyloides enzyme‐linked immunosorbent assay (ELISA) reactivity and (B) arithmetic mean eosinophil counts for group 1 and groups 2 to 4. Dashed lines indicate the upper limit of normal range.


Immigrants from countries in which strongyloidiasis is highly endemic such as Laos and Cambodia4,10,13,24 are at significant risk and so warrant screening, especially if they are to undergo immunosuppressive therapy.14,19 This study examined the ELISA results for S stercoralis and eosinophilia in 22 Laotian immigrants with suspected or proven infection during a period of 3 years after treatment. Our results support a growing body of evidence that serology is a useful diagnostic and management tool in patients in whom parasites are often undetectable in feces.7,9,18,19,25,26 The results also highlight the dilemma facing clinicians in ensuring that S stercoralis parasites have been completely eradicated after treatment.

The usefulness of serology in diagnosing probable strongyloidiasis (that which is not microscopically confirmed), and monitoring the response to therapy, is not well defined due to the lack of a gold standard indicator of infection.7,13,18,19,26 We sought to address this by documenting the serological responses to treatment in a cohort of immigrants living in a temperate area in which S stercoralis is not endemic and reinfection is unlikely. Although there is no way to confirm successful eradication of S stercoralis parasites after treatment, we hypothesized that the consistent and sustained reduction in ELISA to negative levels and resolution of eosinophilia would provide strong circumstantial evidence that a cure had been achieved.

All patients with a strongly reactive ELISA at the time of diagnosis (groups 3 and 4) showed a decline in ELISA reactivity in the first 6 months following treatment, which was accompanied by a decline in eosinophil count, suggesting that these patients were infected with S stercoralis and that parasite numbers were reduced by therapy. However, at 36 months after treatment, one of these patients had died of overwhelming strongyloidiasis after inadvertent treatment with corticosteroids,15 50% (5/10) of the remaining patients had an increased OD compared to their last test, and only one patient had a negative result. Therefore, treatment with albendazole may have failed in up to 50% of patients in groups 3 and 4. This conclusion is supported by other reports of the inadequacy of albendazole for the treatment of strongyloidiasis27–30 and the laboratory confirmation of disseminated strongyloidiasis in one patient.15 It is notable that none of the patients in group 4 achieved levels of ELISA reactivity within the normal range following treatment, raising the question as to whether radical cure was achieved in any of these patients or whether antibody levels in some patients may take many years to completely return to the negative range after successful treatment.18 It is unlikely that unidentified intestinal helminths cross‐reacted in the ELISA in this study causing false‐positive results as most helminths have a relatively short life span, and reinfection would not be expected to occur in our temperate climate.

Patients with low to moderate ELISA reactivity and eosinophilia at baseline (group 2) may have had strongyloidiasis because ELISA reactivity fell in all these patients and eosinophilia had resolved in all but one of these patients by 36 months after treatment. Although ELISA reactivity and eosinophil counts remained relatively unchanged in group 1 patients during the 3‐year period after treatment, consistent with our hypothesis that persistent S stercoralis infection was unlikely in this group, reactivity remained positive in four of the seven patients. Eosinophilia due to strongyloidiasis is notoriously prone to fluctuation and may be absent in at least 20% of patients with microscopically confirmed infection.2,7,19 We therefore urge caution in ruling out strongyloidiasis infection in high‐risk patients who have low to moderate ELISA reactivity without eosinophilia.

The present findings are consistent with previous studies that have shown a reduction in ELISA reactivity following treatment of S stercoralis infections.7,18,25 Kobayashi and colleagues18 examined ELISA responses in patients with positive stool microscopy prior to and 12 to 18 months after treatment with either pyrvinium pamoate or thiabendazole and reported a significant reduction in ELISA values in a large subset of patients. This reduction was highly correlated with a negative post‐treatment stool examination. Similarly, Loutfy and colleagues7 followed antibody responses and eosinophil counts for up to 18 months in patients with a positive stool examination, following treatment with either albendazole or thiabendazole. They reported that ELISA reactivity fell in most patients and that the majority of this response occurred within the first 9 months. Results from the longer follow‐up period in our study suggest that although antibody levels usually fall in the first 6 months after treatment, they may remain positive or even increase in the following 12 to 24 months, raising the possibility of treatment failure.

In summary, this study has shown that an initial reduction in ELISA reactivity does not necessarily indicate eradication of the parasite, the ultimate goal of therapy for Strongyloides infections.2 A reduction in antibody levels may indicate a diminution in the total parasite burden rather than parasite eradication. Recent evidence that agar‐plate culture significantly increases the diagnostic yield in those with strongyloidiasis31,32 may assist in post‐treatment monitoring. Further longitudinal studies, in which adequate therapy is given to ensure a high likelihood of parasite eradication, are required to document performance of the Strongyloides ELISA as a tool for monitoring successful treatment.


The authors wish to express thanks to Sean Tobin, Mao Saykao, Mrs Nuon, Virginia de Crespigny, Mirella Ozols, and Karl Jenkinson for valuable assistance in the conduct of this study and/or preparation of the manuscript. This work was supported by Department of Medicine, The Royal Melbourne Hospital, The University of Melbourne.

Declaration of Interests

The authors state they have no conflicts of interest.


View Abstract