Introduction

A series of vector-borne filarioids belonging to the genera Dirofilaria, Acanthocheilonema, Onchocerca and Cercopithifilaria (Spirurida, Onchocercidae) infect dogs in Europe. Among these, Dirofilaria immitis and Dirofilaria repens cause a severe cardio-pulmonary affection and a mild dermatological condition, respectively (Venco 2007; Albanese et al. 2013). Although both species are regarded as zoonotic agents (Orihel and Eberhard 1998), most human cases in Europe are caused by D. repens (Pampiglione and Rivasi 2000; Otranto et al. 2013). During the last decades, Dirofilaria spp. have expanded their geographical boundaries towards eastern and northeastern countries, in relation to several factors, such as the availability of vector species and climate (Genchi et al. 2009, 2011). However, D. repens seems to be spreading more rapidly than D. immitis (Pantchev et al. 2009a, 2011; Genchi et al. 2011), with autochthonous cases reported in several countries previously regarded as non-endemic, such as in Czech Republic (Svobodová et al. 2006; Dobesová et al. 2007), Germany (Hermosilla et al. 2006; Pantchev et al. 2009b; Sassnau and Genchi 2013), Hungary (Fok et al. 2007), Poland (Sapierzyński et al. 2010; Masny et al. 2011), Slovakia (Miterpáková et al. 2010; Bocková et al. 2013), Ukraine (Hamel et al. 2013) and Austria (Silbermayr et al. 2014). Autochthonous cases of D. immitis infection in dogs have also been suggested based on detection of DNA in whole mosquitoes (Kronefeld et al. 2014) or reported in all the above mentioned countries, except Austria (Svobodová et al. 2006; Miterpáková et al. 2010; Świątalska and Demiaszkiewicz 2012; Hamel et al. 2013; Krämer et al. 2014; Tolnai et al. 2014). However, the detection of filarioid DNA in mosquitoes is not necessarily a proof of stable transmission within a region, as no identification of infectious third-stage larvae was provided yet. Furthermore, in Czech Republic and Poland, the first suggested cases of autochthonous D. immitis infection are questionable, as they relied only on immunological evidence of infection but they were not confirmed by direct (e.g. microfilariae) or molecular methods (species-specific PCR/sequencing) (Svobodová et al. 2006; Świątalska and Demiaszkiewicz 2012). A recent study seems to support this, at least in Poland, where the examination of 1588 canine blood samples (2011–2013) revealed only the presence of D. repens, but not that of D. immitis, in the country (Demiaszkiewicz et al. 2014).

Other species of filarioids affecting dogs (e.g. Acanthocheilonema reconditum, Dipetalonema (syn. Acanthocheilonema) dracunculoides, Cercopithifilaria grassii, Cercopithifilaria bainae and Onchocerca lupi) have also been reported in Europe, but due to their minimal clinical importance (with the exception of O. lupi), they are poorly known (reviewed by Otranto et al. (2013)).

There are several approaches for the diagnosis of filarial infections in dogs. Classical methods applicable for all species with blood-circulating microfilariae include morphological identification (i.e. blood smears or concentration methods, such as Knott’s or filtration test) or histochemical staining of microfilariae. More recently, multiple tools for the highly specific molecular identification of various species of filarioids became available. One disadvantage of all direct methods is that their outcome depends on the presence or the number of microfilariae in the examined sample (Genchi et al. 2007; Pantchev et al. 2011; Latrofa et al. 2012). Other methods (i.e. ELISA and immunocromatographic tests) that can detect circulating antigens of adult female nematodes are currently available only for D. immitis and are recommended as the most sensitive by the American Heartworm Society (2014) because they are also useful in the detection of amicrofilaremic infections. However, cross-reactivity of some commercially available antigen tests for D. immitis with Angiostrongylus vasorum has been recently described and should be taken into consideration in endemic areas where parasites are sympatric (Schnyder and Deplazes 2012; Krämer et al. 2014). Like in most of the countries from Eastern Europe, the current occurrence of filarial infections in dogs from Romania is still unclear. The first extensive epidemiological study was performed in 1933 by Popesco, who described 20 foci of canine filariasis along rivers in the southwest, south and southeast of the country, based on the visualization of microfilariae from blood samples. However, in the report above, morphological details of the microfilariae were not provided. Later on, a nationwide serological screening was performed for D. immitis (Mircean et al. 2012), but data regarding D. repens were only recorded locally, in four counties in the western (Ciocan et al. 2010, 2013), northeastern (Paşca et al. 2008) and southern (Tudor et al. 2013) regions of Romania. Furthermore, four dogs exported from Romania to Germany were positive for microfilariae of D. repens, confirmed by molecular methods (Pantchev et al. 2011). In addition, A. reconditum was also diagnosed in Germany in dogs exported from Romania (Hamel et al. 2012) and recently C. bainae has been reported in a dog from Danube Delta region (Ionică et al. 2014).

The objectives of the present study were to provide a more detailed view on filarial infections in dogs from Romania by assessing the prevalence and diversity of filarioid species infecting dogs from various areas of the country and to compare two different diagnostic tests employed for diagnosing D. immitis.

Materials and methods

Study areas and sampling

From July 2010 to March 2011, 390 blood samples were collected from randomly selected owned dogs from nine counties situated in five regions of Romania, presenting different ecological conditions. Most sampling sites were in rural localities, where dogs were housed outdoors and they generally did not receive regular antiparasitic treatments. With the owners’ consent, samples were collected from the cephalic vein of each dog using a clotting activator S-Monovette syringe (SARSTEDT AG & Co., Germany). After separation, serum was collected into a separate, labelled tube and stored at −20 °C until further processing. The remaining blood clots were kept and stored under the same conditions.

Serological assays

Serum samples were tested for the presence of D. immitis antigen by using an in-clinic SNAP® 4Dx® test (IDEXX Laboratories, Inc., Westbrook, ME, USA), in accordance with the manufacturer’s instructions. This is a rapid assay test system based on enzyme immunoassay technique and has been validated for dogs, having a sensitivity of 99.2 % (Chandrashekar et al. 2010). The D. immitis analyte is derived from antibodies specific to heartworm antigens, which are primarily produced by adult females (Weil 1987).

Molecular assays

Genomic DNA was extracted from the blood clots using a phenol-chloroform method (Maslov et al. 1996; Albrechtová et al. 2011). Approximately 200 μl of clotted blood was dried at 56 °C for 30 min and then suspended in 1.5 ml lysis buffer (0.1 M NaCl, 0.05 M EDTA, 0.01 M Tris, 4.8 % SDS; pH 8) and digested at 56 °C with 20 μl proteinase K (Bioline, UK) for 1 h. After proteins lysis, the mixture was extracted with a 1:1 blend of phenol and chloroform, followed by one extraction with chloroform alone. Each extraction was performed by 1 min of shaking and a 10-min centrifugation step (13.000×g). DNA was precipitated with 96 % ethanol for 15 min and the dried DNA pellet was re-suspended by adding 100 μl of PCR water.

Multiplex PCRs amplifying partial cytochrome c oxidase subunit 1 (cox1) gene regions of different sizes for three filarioid species (D. immitis, 169 bp; D. repens, 479 bp; and A. reconditum, 589 bp) were performed using species-specific forward primers coupled with the reverse primer NTR, following reaction procedures and protocols described in literature (Latrofa et al. 2012). In each set of reactions, a positive control and a sample with no DNA were included in order to test the specificity of the reaction and to assess the presence of contaminants. The positive control sample was obtained by mixing DNA of all three filarioid species, isolated from blood of infected dogs and confirmed through sequencing. PCR products were visualized by electrophoresis in a 2 % agarose gel stained with RedSafe™ 20000x Nucleic Acid Staining Solution (Chembio, UK), and their molecular weight was assessed by comparison to a molecular marker (O’GeneRuler™ 100 bp DNA Ladder, Thermo Fisher Scientific Inc., USA).

Statistical analysis

Data analysis was performed using EpiInfo 7 software (CDC, USA). The frequency of infection, prevalence and its 95 % confidence intervals were established and the differences of prevalence between identified filarioid species and between the two D. immitis diagnostic tests were assessed using chi-square testing. The differences were considered significant if p values were lower than 0.05.

Serological (SNAP®) and molecular (multiplex PCR) methods used for D. immitis detection were evaluated in EpiTools (Sergeant 2014). Agreement between SNAP and PCR was calculated using overall agreement measure and Cohen’s Kappa statistic. As the overall agreement does not differentiate between the agreement on the positives and agreement on the negatives, positive and negative percent agreements were also calculated. A value of k <0 indicates no agreement, between 0 and 0.20 slight agreement, 0.21–0.40 fair agreement, 0.41–0.60 moderate agreement, 0.61–0.80 substantial agreement, and 0.81–1 almost perfect agreement (Landis and Koch 1977).

Results

Molecular diagnosis

Out of 390 sampled dogs, 11.79 % (n = 46) were positive for DNA of at least one filarial species. The overall prevalence of each species was as follows: 6.15 % (95 % confidence interval (CI) 4.07–9.14 %) for D. immitis, 6.92 % (95 % CI 4.70–10.03 %) for D. repens and 2.05 % (95 % CI 0.96–4.16 %) for A. reconditum, with significant local variation (Table 1). The prevalence of A. reconditum was significantly lower (p < 0.001) compared to Dirofilaria spp., but its distribution range was more extended (Fig. 1). Coinfections with D. immitis and D. repens were detected in 23.91 % (n = 11) of positive dogs and those with D. repens and A. reconditum in 4.34 % (n = 2) of the positive dogs.

Table 1 Molecular prevalence of filarial species for each county
Full size table
Fig. 1
figure 1

Distribution of identified filarioid species, established by both employed tests

Full size image

Serology

Meanwhile, 7.18 % (95 % CI 4.91–10.33) of dogs, deriving from three counties, were seropositive for D. immitis antigen at SNAP tests.

Method comparison

Overall, 9.48 % (n = 37) of dogs were positive for D. immitis at least in one of the performed tests. Immunoenzymatic tests have shown a slightly higher prevalence of D. immitis infection in dogs compared to the molecular method (Table 2), but without statistically significant differences (p > 0.5). However, when considering each positive individual (Table 3), the positive percent agreement was in fact 40.54 %. The negative percent agreement was 90.51 % and the overall agreement scored 94.36 %. The k value was of 0.55 (0.38–0.72), indicating a moderate agreement between the two tests. Discordant results consisted in the following: (i) Six samples (16.21 %) were positive only for heartworm antigen, (ii) seven samples (18.91 %) were positive for D. immitis antigen but tested positive for D. repens at molecular methods, and (iii) nine (24.32 %) samples were antigen negative in animals which scored positive for D. immitis at PCR.

Table 2 Seroprevalence and molecular prevalence of D. immitis for each county
Full size table
Table 3 The complete filarial profile of each positive dog
Full size table

Discussion

In Romania, data from the first half of the twentieth Century (Popovici 1916; Popesco 1933, 1935) suggest that southern regions are historically endemic for canine dirofilariasis, although it is unclear which filarial species the authors were referring to. The climate of this country is temperate-continental of transitional type, with four clearly defined seasons varying at regional level according mainly to altitude. In this context, periods when the development of both D. immitis and D. repens can occur are longest (May–October) in the south and southeast, followed by the west and southwest (May–September) and shorter (June–August or September) in the rest of the country (Genchi et al. 2011). The highest prevalence rates of both Dirofilaria species were recorded in the counties that include the Danube’s floodplains, where the climate is suitable for the development of some mosquito species (i.e. Anopheles maculipennis, Culex pipiens) which are confirmed vector species for these filarioids (Nicolescu et al. 2003). Similar to the situation described by Popesco (1933), D. immitis was only identified in proximity of major rivers (i.e. Olt in Braşov county and Danube in Dolj, Teleorman and Tulcea counties) whereas D. repens had a wider distribution range. Indeed, where both species occur in sympatry, the prevalence of D. repens generally exceeded that of D. immitis, as also recorded in other parts of Europe (Genchi et al. 2011). This might be the effect of a protective cross-immunity at individual level, as inferred by the experimental infection of dogs initially infected with D. repens, in which the ability of D. immitis to develop was reduced (Genchi et al. 1995).

The distribution of A. reconditum is herein investigated for the first time in Romania and it seems to occur in a large territory, despite its relatively low prevalence. Transmission of this filarioid is via fleas or lice (Nelson 1962) and requires proximity between the infected and non-infected dogs (Brianti et al. 2012). Compared to Dirofilaria spp., this species appears to be better adapted to dry areas (Constanţa county) and colder climate (Vâlcea county), but so far, its development in the vector in relation to temperature or other climatic factors has not been assessed.

Three types of apparent inconsistencies between serological SNAP tests and PCR diagnosis have been identified in the diagnosis of D. immitis, but their frequency was significantly lower (p < 0.005) than the level of agreement between the two tests.

The most frequently encountered situation (17.30 %, n = 9) was that animals negative for D. immitis at SNAP test were positive for D. immitis DNA, which suggests the presence of microfilariae or soluble genomic DNA in the blood at sampling time. This may be due to a low number of adult worms, e.g. one to two gravid females, previous adulticidal treatment or delayed antigenaemia based on low worm burdens and chemoprophylaxis (Courtney and Zeng 2001; Nelson et al. 2005; Pantchev et al. 2011). Furthermore, given that D. immitis microfilariae have a life span of up to 2.5 years (Abraham 1988), they could persist after the natural death of adult females. Moreover, recent data has revealed that heating the serum samples before laboratory processing renders more sensitivity to the SNAP test, suggesting the existence of inhibitors of a yet unknown nature in the serum (Little et al. 2014; Velasquez et al. 2014).

Interestingly, 13.46 % (n = 7) of samples were positive for D. immitis antigen while molecular methods identified only D. repens. Since the possibility of cross-reaction between the two species was excluded by Pantchev et al. (2009b, 2011), this finding may suggest the occurrence of a patent D. repens infection associated with an occult D. immitis infection, revealing an interesting pattern that deserves further investigation. Since the actual relationship between the two Dirofilaria species has only been partially studied (Genchi et al. 1995), these results could represent the outcome of a potential inhibition of D. immitis microfilarial production by the presence of D. repens and the host immune responsiveness. In addition, a small number of D. immitis microfilariae present in the peripheral blood could fall under the detection limit of the employed PCR method (i.e. 26 microfilariae/ml; Latrofa et al. 2012). As microfilaremia fluctuates during the day, false negative results may emerge according to the sampling time. For both species, the periodicity of microfilariae has been assessed and it seems to vary not only with external factors like the feeding behaviour of vector species (indicating the optimum sampling time is during the evening and night), but also with internal (host-related) factors, like the blood oxygen pressure, which decreases while the animal is sleeping, determining a rise in microfilaremia (Hawking 1956; Aoki et al. 2011; Di Cesare et al. 2013).

Some samples (11.53 %, n = 6) were positive only for D. immitis antigen and negative for all filarioid species by PCR, which may indicate occult (amicrofilaremic) infections, such as in the case of prepatency period, unisexual infection, drug-induced sterility of adults, or immune-mediated clearance of microfilariae (Rawlings et al. 1982). Another potential explanation would be a cross-reaction with antigen of the “French” heartworm A. vasorum (Schnyder and Deplazes 2012). This parasite has not been reported in Romania so far, but models show that parts of the country may be included in its distribution range (Morgan et al. 2009). Nowadays, a revised version of the test system used for the current study, SNAP® 4Dx® Plus (IDEXX Laboratories, Inc., Westbrook, ME, USA), which does not show any cross-reactivity between D. immitis and A. vasorum (Schnyder and Deplazes 2012), and a specific rapid A. vasorum device (Schnyder et al. 2014) that was not on the market when testing for the present study was performed, are commercially available.

Failure of serological tests in detecting patent infections may have serious implications in the spreading of the disease and in the clinical outcome, so they should not be used as the only screening method in epidemiological studies. Since molecular methods offer the possibility to identify more filarioid species, we regard them as a necessary additional screening tool for surveillance, also taking into consideration the zoonotic potential of D. repens.

Conclusion

The present study shows that in Romania, Dirofilaria species are commonly present in the south and southeast of the country The current study is the first to provide a more extensive overview on the prevalence and geographical distribution of A. reconditum in dogs from Romania.