Results and discussion

The overall seroprevalence of strongyloidiasis in our sample of patients with eosinophilia was 9.31% (131 positive samples out of 1407 tested). In seropositive individuals, the total number of positive microscopy findings following stool analysis by the MIFC method was 34 out of 131 (or 25.95% when presented as a fraction). There were no statistically significant differences in either seropositivity (χ² = 6.42; P = 0.093) or percentage positive microscopy rates (χ² = 3.09; P = 0.379) when different study years are compared (table 1).

Table 1. Percentage positivity rates for Strongyloides stercoralis with the use of serology and confirmatory microscopy of samples concentrated by the merthiolate–iodine–formaldehyde concentration method in accordance with study years (note: microscopy was pursued only in seropositive individuals) (P-value < 0.05 was considered as statistically significant; Chi-square test).

Among 1407 individuals whose samples were included in the study, there were 736 male and 671 female patients. No statistically significant differences in seropositivity by gender have been demonstrated, and the percentage rates were quite comparable at 9.24% for women and 9.38% for men (χ² = 0.008; P = 0.931). Moreover, even though we observed a higher percentage of positive microscopy rates following the MIFC method in seropositive individuals when men were compared to women (30.43% vs. 20.97%, respectively), this difference was also not statistically significant (χ² = 1.52; P = 0.217).

The age of our study participants ranged from one to 96 years (mean: 43.7, median: 43, mode: 34, interquartile range: 39). When analysing different age groups, we observed a notably higher seroprevalence rate in those older than 60 years of age (13.10%) when compared to those between 31 and 60 years of age (8.75%) and those 30 years and younger (6.57%) (χ² = 10.7; P = 0.005). Further fragmentation of age groups also showed statistically significant differences, with most serologically positive results observed in those older than 70 years of age (table 2), highlighting in turn the positive correlation of seropositivity and age.

Table 2. Percentage positivity rates for Strongyloides stercoralis with the use of serology and confirmatory microscopy of samples concentrated by the merthiolate–iodine–formaldehyde concentration method in accordance with age groups (note: microscopy was pursued only in seropositive individuals) (*P-value < 0.05 was considered as statistically significant; Chi-square test).

On the other hand, there were no statistically significant differences in regards to age for MIFC microscopy positivity rates (χ² = 0.901; P = 0.924) (table 2). Of note, the positive MIFC microscopy rate was higher in those older than 50 years of age in comparison to those younger than 50 years of age (28.17% vs. 23.33%), which are two study sub-populations of similar size when the number of analysed samples is concerned.

To our knowledge, this is a first study from Croatia that aimed to assess the frequency of strongyloidiasis in Croatian patients with eosinophilia. Our results reveal substantial seroprevalence rates in both women and men and a rather high infection burden in older individuals, but also additionally highlight how the MIFC method is largely insufficient to capture all S. stercoralis infections. The fraction of positive microscopy findings (following the use of the MIFC method) in seropositive individuals in our study (i.e. 25.95%) is highly comparable to a recent study by Balen Topić et al. (Reference Balen Topić, Marjanović, Tomasović and Sviben2021) from Croatia where it was shown how the MIFC method revealed only 26.2% of cases that were positive when more sensitive diagnostic techniques have been used.

The latter is important as the MIFC concentration procedure is still predominantly utilized as a standard diagnostic tool in most microbiology laboratories in Croatia and other neighbouring countries; however, it is actually insufficiently sensitive, taking into account the importance and increasing recognition of this disease. Studies have shown how the Baermann concentration technique and agar plate culture have a much better diagnostic yield, but the number of stool samples, faecal amount and/or faecal dilution also affect sensitivity and detection power (Hailu et al., Reference Hailu, Amor, Nibret, Munshea, Anegagrie, Flores-Chavez, Tang, Saugar and Benito2022). Importantly, a meta-analysis by Campo Polanco et al. (Reference Campo Polanco, Gutiérrez and Cardona Arias2014) that evaluated conventional parasitological methods for the diagnosis of S. stercoralis found the highest sensitivity for the agar plate method (89%), followed by the Baermann technique (72%), formalin–ether concentration technique (which is akin to the MIFC method) (48%) and direct wet smear (21%).

Furthermore, the use of PCR may prevent misidentification of morphologically kindred helminth species (Becker et al., Reference Becker, Piraisoody and Kramme2015), and sometimes even the analysis of aspirates and biopsy samples from upper gastrointestinal endoscopy and/or bronchoscopy is endorsed (Balen Topić et al., Reference Balen Topić, Marjanović, Tomasović and Sviben2021). For both epidemiological studies and quotidian clinical practice, serology is considered a diagnostic mainstay; however, if the accuracy of serology is evaluated against a faecal-based reference standard, discordant results (i.e. serology-positive and faecal-negative results) would make classification endeavours impossible. This issue has been bypassed in several ways, as described by Buonfrate et al. (Reference Buonfrate, Formenti, Perandin and Bisoffi2015).

That said, there is no official single reference standard for diagnosing strongyloidiasis, thus a combination of methods is the most feasible approach. This is particularly pertinent when corticosteroid therapy is prescribed, as previously undetected and/or subclinical Strongyloides infestation can develop into hyperinfection syndrome during treatment and subsequently disseminate, leading to very high mortality rates (Ahmed et al., Reference Ahmed, El-Moselhy, El-Moammaly and El-Shewy2019). Another important group where serological evaluation should be routinely sought are transplantation candidates (Toledo et al., Reference Toledo, Corral and Meisel2019), although reduced sensitivity of serology in immunocompromised individuals should be taken into account. A study by Winnicki et al. (Reference Winnicki, Eder, Mazal, Mayer, Sengölge and Wagner2018) suggests that kidney transplantation programmes in Central Europe should entail recipient and donor screening as Strongyloides may be underestimated.

In other patient groups, especially those that are not critically ill, a stepwise approach can be a good way to increase the diagnostic yield – with MIFC as the first step, and then (if the result is negative), serology or even molecular diagnostics may be pursued in some instances (if available) (Balen Topić et al., Reference Balen Topić, Marjanović, Tomasović and Sviben2021). Underlying patient conditions also have to be taken into account. A study from the United Kingdom showed a significant association of strongyloidiasis with diabetes and eosinophilia (McGuire et al., Reference McGuire, Welch and Melzer2019). Moreover, a steadfast parasitological diagnosis of S. stercoralis infection in alcoholic patients necessitates repeated examination by at least two parasitological methods – including agar plate culture as a result of its higher sensitivity (Silva et al., Reference Silva, Inês and Souza2019).

Our study is a direct contribution to a very scarce body of country-level epidemiological data on S. stercoralis in Europe, which significantly influences the accuracy of various estimation endeavours – the most recent one showing S. stercoralis prevalence of 2.8% and 26.1 million infected individuals in the World Health Organization European region (Buonfrate et al., Reference Buonfrate, Bisanzio and Giorli2020). In a study from San Marino there was a 4.8% seroprevalence rate and one additional case was detected with the use of gastric biopsy (Cappella et al., Reference Cappella, Piscaglia, Cadioli, Manoni, Silva and Buonfrate2019). A recent systematic review of endemic cases in Spain showed a significant concentration of infected individuals in the province of Valencia, with a high predominance of male patients; this is in contrast with our study where there was no statistically significant difference when sexes have been compared (Barroso et al., Reference Barroso, Salvador, Sánchez-Montalvá, Bosch-Nicolau and Molina2019).

Co-infections and occupational information are also relevant pieces of the puzzle, as studies have shown a higher exposure to S. stercoralis larvae among patients with leptospirosis and those working in agriculture, respectively (Varzegar et al., Reference Varzegar, Bayani, Kalantari, Nasiri-Kenari, Amini Navaie, Mollalo and Rostami2021). Some recent studies highlight the need to implement stringent screening due to high seroprevalence rates in migrant populations coming to Europe, highlighting serology as an optimal approach (Requena-Méndez et al., Reference Requena-Méndez, Salas-Coronas and Salvador2020). Furthermore, as a result of the characteristic auto-infective cycle of the parasite, S. stercoralis can persist even though other helminth infections may have disappeared in this region, as was described in Italy (Buonfrate et al., Reference Buonfrate, Baldissera and Abrescia2016). These are undoubtedly important epidemiological concerns that have to be taken into account.

Global age-related findings imply that children are generally not at a higher risk for S. stercoralis infection; nonetheless, behavioural factors might have a significant influence here. Based on the study conducted in Spain, Bustamante et al. (Reference Bustamante, Pérez-Muñoz, Sainz, Lopez-Hortelano, Montero-Vega and Mellado2021) recently argued how strongyloidiasis should always be considered as a differential diagnosis in children presenting with eosinophilia. In our study there were 1.36% patients below 18 years of age with eosinophilia that were seropositive, and 20% of them had a positive microscopy finding following the use of the MIFC method, which adds another layer of evidence that there is a genuine autochthonous transmission.

The link to eosinophilia also warrants a separate discussion, as the literature shows that it might be a much more frequent finding in comparison to other chronic intestinal parasitic infections (Salvador et al., Reference Salvador, Sulleiro, Sánchez-Montalvá, Saugar, Rodríguez, Pahissa and Molina2014; Krolewiecki & Nutman, Reference Krolewiecki and Nutman2019). In that regard, eosinophilia can indeed be viewed as a potentially valuable marker for screening asymptomatic individuals suspected to be infected with S. stercoralis. Nevertheless, it has to be emphasized that eosinophilia in chronic strongyloidiasis might be intermittent and some reports on strongyloidiasis have documented eosinophilia in 57–63% of cases (Requena-Méndez et al., Reference Requena-Méndez, Chiodini, Bisoffi, Buonfrate, Gotuzzo and Muñoz2013); furthermore, it is not a reliable predictor of hyperinfection and it may not be present during the treatment with immuno-suppressant therapy (Greaves et al., Reference Greaves, Coggle, Pollard, Aliyu and Moore2013).

Our study has several important limitations. First, we did not exclude the possibility of infection by some other parasite, as well as other potential causes of eosinophilia; hence, it was impossible to ascertain that S. stercoralis was in fact the actual (or sole) cause of eosinophilia in our positive patients (also taking into account the fact that eosinophilia can be an insensitive marker of infection). Second, as specimens have been received through a reference laboratory, we only had basic patient demographic and clinical information at our disposal, without details regarding their medical history, clinical presentation, the course of eosinophilia or additional laboratory results – hampering other potentially valuable analyses within the epidemiological context. Third, the serological method used prevented us from separating the acute or latent phase of strongyloidiasis in our studied population, and there is a possibility of false-positive reactions as a result of cross-reactivity with other parasites (e.g. Ascaris lumbricoides or Toxocara spp.). Likewise, detailed analytical specificity studies were not conducted, even though serological cross-reactivity among the anti-Strongyloides IgG assays has been observed in individuals who previously had filarial infections. The use of a more sensitive stool detection method could result in different percentage positive rates for confirmatory microscopy. We are also aware that sensitivity can be lower in immunocompromised individuals; however, a minimal influence on the overall results is expected here, taking into account a high number of screened individuals. There are also several types of selection biases pertinent for our study: eosinophilia as a principal inclusion criterion (even though this is absent in a significant proportion of individuals with chronic strongyloidiasis); stool testing conducted only on serologically positive individuals (considering the reported sensitivity of the serological test of approximately 90%); as well as the use of MIFC as a stool test with a reported sensitivity of approximately 48% compared to a composite reference standard.

Notwithstanding the aforementioned limitations, by providing important, country-level data for the first time, we can conclude that this study reinforces the notion of endemicity and strongyloidiasis transmission in Croatia. Although the disease incidence is generally low, patients on immunosuppressive drugs and organ donors from Croatia should be included in the screening programme for S. stercoralis infection – with the caveat that the sensitivity may be reduced in such circumstances, potentially necessitating additional methodologies. As the occurrence of this parasite in the stool can be intermittent and in low concentrations (particularly in chronically infected asymptomatic patients), corroborated by only a quarter of positive microscopy findings in our seropositive patients, serology can be viewed as the preferred diagnostic approach, emphasizing the need for combining different methods whenever possible.