The most common and pathogenic nematode parasites in horses originate from the family Strongylidae and the subfamilies Cyathostominae and Strongylinae, respectively. They differ particularly concerning size and shape of the buccal capsule (1). The Cyathostominae (cyathostomins or small strongyles) show the highest prevalence among all helminths of horses. They do not perform tissue migration during their development in the host although they have intramucosal stages that might also become hypobiotic (2–8). In contrast, the most important members of the Strongylinae (large strongyles), which belong to the genus Strongylus, show complex species-specific tissue migrations through different abdominal organs over several months. During these migrations, they can cause severe tissue damage, clinical signs, and can lead to life threatening complications (9, 10).

Previous studies revealed high prevalence of strongyles on German horse farms with 91.6% by Wirtherle et al. (11) in Lower Saxony, 98.7% by Fritzen et al. (12) in North Rhine-Westphalia and 98.4% by Hinney et al. (13) in Brandenburg on the farm level. A modified McMaster method was used in all these studies. Additional coproscopic techniques (combined sedimentation/flotation technique, perianal tape test, sedimentation, and Baermann-Wetzel method) were only performed in the latter study. Kaspar et al. (14) reported a prevalence of 55.3% on the individual horse level in a study testing horses from 91 stables across Germany using a modified McMaster method and combined sedimentation/flotation. In another study using the modified McMaster method, 44.6% of the equines from 192 horse farms across Germany were found to shed strongyle eggs (15). To date, cyathostomins are considered to contribute to the vast majority of the egg shedding, since the prevalence of large strongyles was found to have dropped to very low levels following the introduction of anthelmintics with efficacy against migratory stages of these parasites in the early 1980's (16, 17). Table 1 shows the prevalence of Strongylus vulgaris as determined in earlier studies conducted in Germany. Using larval culture and subsequent morphological differentiation of the third stage larvae (L3), prevalences of 0.1–1.3% were observed (13–15, 18, 23).

The prepatent period of S. vulgaris ranges from six to seven months (24). After oral ingestion of the infective third larval stage (L3), which develops on pasture, L3 undergoes exsheathment and penetrates the mucosa of the caecum and colon (25). The life cycle of S. vulgaris comprises an extensive parenteral larval migration. The L3 moult in the mucosa of the large intestine to L4 and starting 2 weeks post-infection begin to migrate through the intestinal arteries and the cranial mesenteric artery. They migrate both on and in the intima of the blood vessels. After about 3 months, another moult to the preadult stage occurs, and another 4–6 weeks later the worms move back with the blood stream towards the intestine (26). The most frequent clinical signs of infection with S. vulgaris are fever, lethargy, weight loss, and colic (26–28). The extensive parenteral migration of the larvae in the cranial mesenteric artery and its branches causes endothelial damage, resulting in inflammatory reactions. The endothelial damages lead to the development of thrombi from which emboli result, which end up in the small and large intestinal wall leading to haemorrhagic infarctions. Thickening of the arterial walls occurs (24), leading to impaired blood flow and sometimes severe damage of the arterial wall due to passive dilatation which has been reported to also result in sponge-like so-called “verminous aneurysms” (29–34). In more recent reports clinical infections with S. vulgaris have been associated with the appearance of non-strangulating intestinal infarction, leading to peritonitis and mild to severe colic (27, 35).

The larvae of Strongylus edentatus migrate to the liver through the intestinal veins. An infection is rarely reported to be associated with clinical signs (36, 37). Like S. edentatus, the larvae of Strongylus equinus also migrate through the liver, but they also pass through the pancreas and cause inflammation in both organs (38). Strongylus asini is described to occur in zebras and donkeys. It is reported to cause an extensive liver pathology. Horses are much less susceptible to S. asini than other equine species (39).

The prepatency of the cyatostomins ranges from 1.5 to 3 months (40–42). However, cyathostomin larvae can remain encysted in the caecal or colonic mucosa for 2 years or even longer when the maturation of the early L3 is arrested (7, 43, 44). Cyathostomin infections are associated with the occurrence of milder symptoms such as weight loss. However, synchronised egress of hypobiotic larvae from the mucosa can cause larval cyathostominosis that may lead to severe clinical signs such as weight loss, protein-losing diarrhoea and colic, and even death (43, 45, 46).

Detection of an equine infection with nematodes is often carried out using a coproscopic method. However, it is not possible to reliably distinguish the eggs of Cyathostominae and Strongylinae based on their morphology. Instead a larval culture is required for this purpose, since the L3 can be distinguished microscopically between the two and also to the species level concerning Strongylus spp. using details such as the number of mid gut cells (47).

A real-time PCR (48), conventional PCR, and other PCR-based methods (48–51) have been described to identify patent infections with S. vulgaris using eggs or larval stages. It was reported that significantly higher numbers of horses were tested S. vulgaris positive by real-time PCR and conventional PCR than by morphological examination of L3 from larval cultures (14, 52). Serological detection of infections with cyathostomin species (53) or S. vulgaris (54) have been described to be possible using enzyme-linked immunosorbent assays (ELISAs).

Currently, three broad-spectrum anthelmintic classes are commercially available by prescription for nematode control in horses in Germany: Tetrahydropyrimidines (pyrantel pamoate), macrocyclic lactones (MLs, ivermectin, moxidectin), and benzimidazoles (BZs, fenbendazole). They have been used for decades to treat and control equine nematode infections. Encysted, hypobiotic cyathostomin larvae are known to be rather insensitive to currently available anthelmintics but can be affected by treatment with moxidectin (with moderate efficacy) (16) as well as a five-day treatment with fenbendazole, the latter provided the present cyathostomin population has not developed BZ resistance (55, 56).

Anthelmintic resistance (AR) in equine nematodes is widespread (4, 57–59). Concerning strongyles it is primarily the BZ drug class against which cyathostomin populations have developed resistance, but also resistance to pyrantel is prevalent (57, 60–66). Molecular analysis of AR associated genetic changes, for example concerning the respective drug target, promises to allow the detection of AR development at earlier stages than phenotypic tests allow (67). Benzimidazoles inhibit the polymerization of α- and β-tubulin dimers into microtubules (68, 69). Disruption of microtubule polymerisation is supposed to affect the energy metabolism and inhibit the fumarate reductase and the glucose uptake in some nematode species, which leads to the death of the worms (70). Benzimidazole resistance in strongyle nematodes is particularly associated with single-nucleotide polymorphisms (SNPs) in the isotype 1 β tubulin gene. Exchanges in codons 167, 198, or 200 have been described to be associated with BZ resistance (71–78). It has been recorded that the SNP at codon 167 was predominantly associated with BZ resistance in cyathostomins (79).

The increasing emergence of AR of cyathostomins worldwide is a substantial concern in terms of recommendations of appropriate parasite control schemes (80). A surveillance-based approach using faecal egg count analysis (FEC) before anthelmintic treatment has been introduced in several European countries such as Denmark, Sweden, Finland, the Netherlands, and Italy (81). A prescription-only sale of anthelmintic drugs for horses has been implemented with the aim to reduce the risk of AR development. However, no association between presence of S. vulgaris and the magnitude of the FEC has been found (82, 83). A reappearance of S. vulgaris has been reported in Denmark and Sweden upon the introduction of prescription-only policies (82, 84).

The aim of this article is to provide: (i) an overview about the distribution of nematodes in horses in Berlin and Brandenburg, Germany and (ii) an insight into the occurrence of large strongyles with a focus on S. vulgaris. Additional aims were to describe (iii) a real-time PCR with subsequent high-resolution melt (HRM) analysis for species differentiation of large strongyles other than S. vulgaris and (iv) to develop pyrosequencing assays for S. vulgaris to determine allele frequencies in the BZ resistance associated codons 167, 198, and 200 of the β-tubulin isotype 1 gene.

In the present study, prevalence of strongyle eggs on the individual horse as well as on the farm level were within the range observed in other studies from Europe before. High individual horse prevalences ranging up to 100% have been reported (13, 95, 96), although two studies conducted in Germany found lower prevalence on individual level: prevalences of 44.6% (15) and 55.3% (14) were determined. Concerning the true strongyle infection prevalence, the individual horse coproscopic prevalences, herein 66.7%, must always be considered against the background that only patent infections can be detected using coproscopy. Moreover, some patent infections will most likely remain undetected due to variation in egg shedding between droppings (97). More importantly, encysted larvae have been shown to represent the major part of the total cyathostomin burdens (95).

The results of the S. vulgaris real-time PCR, in which 0.83% of the samples were positive, are consistent with findings of previous studies that also reported a prevalence of <2% (horse level) using larval culture (14, 15) and PCR (14). The PCR HRM assay to detect S. edentatus and S. equinus identified the presence of S. edentatus in 10 horse samples, whereas no amplification of S. asini and S. equinus was observed. This is in line with the reported prevalences of large strongyles in other publications, since S. vulgaris and S. edentatus were reported to occur more frequently in horses than S. equinus (21, 98). Individual horse prevalences of 0–44% for S. edentatus and 0–10.9% for S. equinus have been reported in previous studies conducted in Germany (19–22). For technical and convenience reasons, primers were designed in a way to be also able to also amplify S. asini, since the complete group of Strongylus species would be of relevance in possible subsequent studies. Infection of horses with S. asini has to the knowledge of the authors not been reported from naturally infected horses and the only reported experimental infection of a single horse led to lower pathology caused by migrating larvae than in a zebra and a donkey infected in parallel (39). Clear evidence whether S. asini can lead to patent infection in horses is missing.

Table 4 provides details of the FEC of the horses that were tested S. edentatus- and/or S. vulgaris-positive with real-time PCRs. It is noticeable that all positive horse samples were collected between May and July 2017. No horse was tested positive for large strongyles that had been sampled in the later period (between August 2017 and January 2018). This can be explained by the life cycle of S. vulgaris with larvae undergoing migration over winter and sexually mature adults start shedding eggs in spring (99, 100). Later in the year, the practise of regular anthelmintic treatment during the grazing period could eliminate the adult S. vulgaris leading to absence of patent infections despite presence of the parasite on the farm. The majority of the study population received anthelmintic treatment at regular intervals (Table 2). For the four samples that were tested positive for S. vulgaris, no polymorphisms associated with BZ resistance detected in other strongyle nematodes were found herein neither using pyro- nor Sanger sequencing in the S. vulgaris-specific amplicons. The newly developed pyrosequencing assays can be potentially used in future epidemiological surveys to determine whether BZ resistance associated genetic changes will evolve in large strongyles similarly as it has been documented in several ruminant helminth species.

Since FECs are in most cases dominated by eggs of cyathostomins and cannot be discriminated from eggs of large strongyles in a FECRT, BZ resistance of large strongyles would not be recognised against the background of widespread BZ resistance in cyathostomins. Molecular approaches to detect large strongyles in post-treatment samples such as the real-time PCRs used here and assays to directly quantify molecular AR markers such as the novel pyrosequencing assay for S. vulgaris promise to help to perform AR monitoring for these highly pathogenic equine parasites.

The number of horses that were tested S. vulgaris positive in the ELISA was much higher than with the PCR. According to Andersen, et al. (54), there is a statistical correlation between a higher anti-rSvSXP titre and the number of migrating larvae in the mesenteric arteries. The rSvSXP-ELISA was introduced as an improved method to detect early stages of S. vulgaris during prepatency of an infection. The horses tested positive for S. vulgaris by real-time PCR had at least an ODR of 16.9% in the ELISA. Thus, all of these horses had ODRs above the original cut-off ODR of 13.5% with a sensitivity of 0.66 and a specificity of 0.82 (54).

The cut-off values for the rSvSXP-ELISA applied in the present study were 25 and 30%, which is a more conservative approach decreasing the number of false positives but also increasing the number of false negatives. High S. vulgaris seroprevalences were also reported in other studies (91, 101), but are unexpected in light of surveys employing direct pathogen detection methods (14, 15). The high number of serologically positive horses compared to the four horses tested S. vulgaris positive by real-time PCR may partly be due to the fact that early stages in the prepatency of the infection can be detected only with the ELISA. Moreover, antibody levels only decrease to background after 5 months post infection (102). Therefore, seropositivity is probably due to frequent exposure with infective larvae that start migration while patency only develops in a few horses since the developmental cycle is disrupted by anthelmintic treatment in most of the exposed animals. However, the small number of horses in which adult S. vulgaris develop and produce eggs is obviously sufficient to contaminate the pasture leading to high seroprevalence in the respective horse population. Another important issue that needs to be taken into consideration is a potential cross-reactivity of rSvSXP with the orthologous protein from other Strongylus species such as S. edentatus and the cyathostomins, which appear to be phylogenetically closely related but have not been characterised for most cyathostomin species (54).

The multivariate risk factor analysis (logistic regression) revealed that the odds to shed strongyle nematode eggs were significantly increased with a longer period between the last anthelmintic treatment and sampling. This is expectable since chances to be reinfected increase over time. In addition, the analysis indicated that the odds of shedding strongyle type eggs increases with an Anoplocephala spp. positive serological ELISA but this effect was not significant. This variable probably does not directly influence the odds and represents a confounder since both parasite groups share transmission via grazing. Therefore, it is also reasonable that limited/no access to pasture was protective, because it means lower exposure to infectious larval stages. Age as protective factor is consistent with the perception of young horses showing a higher incidence of patent strongyle infections due to age-dependent immunity developing through continuous exposure (103, 104). With ivermectin and particularly strong with moxidectin, the anthelmintic used for the last treatment had a protective effect compared to the reference fenbendazole. Treatments with ivermectin plus praziquantel, pyrantel, and doramectin treatments were also protective but the effects were not significant. For ivermectin and ivermectin plus praziquantel the odds ratios were similar but the effect for the latter was not significant, possibly since only a very small number of horses was treated with the combination. Doramectin was used only for a few horses from a single farm and it must be clearly stated that doramectin is not licenced as a dewormer of horses and that the ruminant product was used. Resistance to BZs and pyrantel is well-known to occur in many European countries (61, 63, 105) including Germany (64) at high prevalence. This might lead to treatment failures (faecal egg count reduction below 90–95%) (57) or to shorter egg reappearance periods (106). Resistance to ivermectin and particularly moxidectin in cyathostomins have only rarely been reported on a global scale but so far not in Germany.

The logistic regression analysis of the data on S. vulgaris seroprevalence identified only a few variables that had significant effects on the odds to be positive, i.e., “pasture access”, “treatment scheme” and the age of the horses. Regarding “pasture access”, only “sometimes pasture access” was significantly protective in comparison to “always pasture access” while “no pasture access” was surprisingly not protective. The odds ratio was close to 1 but the 95% CI was also very wide which is due to the fact that only 16 horses (3.3%) were assigned to this category. Increasing age was again slightly protective, which can be explained by immunity slowly developing over time. In contrast to the analysis on the odds to be positive for shedding strongyle eggs, the treatment scheme had a significant effect on the odds to be positive for antibodies against the S. vulgaris larval antigen. There was a clear tendency that higher treatment frequencies were associated with lower odds to have antibodies against rSvSXP. However, only the difference between four treatments per year and selective treatment was significant. Using morphological analysis of L3 from larval cultures, Nielsen et al. (84) have previously reported that selective deworming strategies resulted in roughly doubled prevalences of S. vulgaris on the individual animal and farm level in Denmark. It is also remarkable that in contrast to the analysis of data on positivity of faecal samples, the drug used for the last treatment was excluded from the final regression model for seropositivity for the larval S. vulgaris antigen since it led to an increased AIC. This substantiates the idea that the odds to be positive for antibodies against S. vulgaris are more dependent on the contamination of the pasture (and thus the treatment history of the farm) and not on the treatment history of the individual horse since for immune reactions and seropositivity the exposure is relevant and not the risk to develop a patent infection. Out of the 42 horses for which the owners reported “selective treatment”, 41 were from only 4 farms on which all horses were assigned to the “selective treatment” group. This confirms that the “selective treatment” scheme was highly clustered on the farm level.

PCR assays identified S. vulgaris and S. edentatus at only low frequencies, which is in contrast to the high S. vulgaris seroprevalence observed herein. The cause for the very high prevalence of antibodies against rSvSXP remains unclear and should be further examined to exclude any confounders such as cross reactivity. The fact that an important risk factor for being positive for antibodies against a larval S. vulgaris antigen was a selective treatment scheme based on prior diagnosis, emphasises the risk of such treatment schemes, if they do not include regular analysis of eggs or larvae or antibodies for the presence of Strongylus spp. on a farm.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Landesamt für Gesundheit und Soziales, Berlin, registration no: Reg 0059/17. Written informed consent was obtained from the owners for the participation of their animals in this study.

LJ, JK, EB, JB, MN, and GS-H: study concept and design. LJ: sample collection and Mini-FLOTAC and PCR analyses. NH: involved in establishment of the Strongylus spp. HRM PCR. GS-H funding acquisition and project management. LJ and JK: data analyses and drafting the first version of the manuscript. All authors contributed to the article and approved the final manuscript version.

This work was funded by Virbac in a research collaboration with Freie Universität Berlin (contract number 2014000271).

JB is employed by Virbac Tierarzneimittel GMBH and EB by Virbac France. GS-H declares that he has repeatedly acted as consultant for veterinary pharmaceutical and diagnostic companies and has previous and ongoing research collaborations with various companies. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.