Cats and Toxoplasma gondii: A systematic review and meta-analysis in Iran

Abstract Toxoplasma gondii is a cosmopolitan zoonotic intracellular coccidian of the phylum Apicomplexa infecting warm-blooded animals and human beings. This protozoan causes a significant public health problem in humans and imposes considerable economic losses and damages to husbandry industries. The final host, cats, accounts for all of these significant burdens. Hence the present study was designed to analyse and review the overall prevalence rate of T. gondii infection in cats in Iran for the first time. In the present study data collection (published and unpublished papers, abstracts of proceedings of national parasitology congresses and dissertations) was systematically undertaken on electronic databases including PubMed, Google Scholar, Ebsco, Science Direct, Scopus, Magiran, Irandoc, IranMedex and Scientific Information Database. A total of 21 studies from 1975 to 2013 reporting prevalence of Toxoplasma infection in cats from different areas in Iran met the eligibility criteria. The pooled proportion of toxoplasmosis using the random-effect model amongst cats was estimated at 33.6% (95% confidence interval [CI] 22.05–46.41). The prevalence rate of cat toxoplasmosis in various regions of Iran ranged from 1.2% to 89.2%. Firstly, this study establishes a crude prevalence rate of T. gondii infection in cats. Secondly, it discusses the role of significant risk factors including sex, age and being either household or stray cats, in the epidemiology of the disease. Furthermore, the current study determines gaps and drawbacks in the prior studies that are useful to keep in mind to assist in designing more accurate investigations in future.


Introduction
Toxoplasma gondii, the causative agent of toxoplasmosis, is an obligate intracellular parasite which belongs to the phylum Apicomplexa that infects all species of warm-blooded animals (Flegr et al. 2003). In humans it is considered to be one of the most common parasites, based on serological investigations that estimate that up to a third of the world's population has been exposed to this widespread zoonotic agent. The overall seroprevalence rate of toxoplasmosis amongst the general population in Iran is 39.3% (95% confidence interval [CI] 33.0% -45.7%) (Bahrami et al. 2011;Daryani et al. 2014;Dubey & Jones 2008;Sharif et al. 2007).
Even though the majority of toxoplasmosis cases in immune-competent individuals are either asymptomatic or mild, first exposure to T. gondii during pregnancy can lead to transplacental transmission to the embryo, with serious pathological signs including hydrocephalus, microcephaly, blindness, abortion and death of the foetus (Dunn et al. 1999;Havelaar, Kemmeren & Kortbeek 2007). In addition, T. gondii is considered to be an opportunistic and life-threatening parasite in immune-compromised groups, encompassing those with HIV and AIDS, cancer and organ transplant recipients who receive immunosuppressive drugs (Tenter, Heckeroth & Weiss 2000).
Felids play a pivotal role for T. gondii as definitive hosts, and interestingly are known as the only final hosts that produce oocysts in their faeces, contaminating soil, food and water (Cenci-Goga et al. 2011;Dubey 2004). Even though the final host excretes oocysts for a short period of time (only 1-2 weeks), millions of oocysts may be excreted. Oocysts can survive in the environment for several months and are noticeably resistant to freezing, drying and disinfectants, whereas they are not heat-resistant and are destroyed at 70 °C for 10 min (Dubey & Beattie 1988;Dubey & Jones 2008).
Infection both in the definitive and the intermediate host usually occurs either through ingestion of infected tissue cysts or oocysts. Felids are more likely to shed oocysts followed by ingestion Toxoplasma gondii is a cosmopolitan zoonotic intracellular coccidian of the phylum Apicomplexa infecting warm-blooded animals and human beings. This protozoan causes a significant public health problem in humans and imposes considerable economic losses and damages to husbandry industries. The final host, cats, accounts for all of these significant burdens. Hence the present study was designed to analyse and review the overall prevalence rate of T. gondii infection in cats in Iran for the first time. In the present study data collection (published and unpublished papers, abstracts of proceedings of national parasitology congresses and dissertations) was systematically undertaken on electronic databases including PubMed, Google Scholar, Ebsco, Science Direct, Scopus, Magiran, Irandoc, IranMedex and Scientific Information Database. A total of 21 studies from 1975 to 2013 reporting prevalence of Toxoplasma infection in cats from different areas in Iran met the eligibility criteria. The pooled proportion of toxoplasmosis using the random-effect model amongst cats was estimated at 33.6% (95% confidence interval . The prevalence rate of cat toxoplasmosis in various regions of Iran ranged from 1.2% to 89.2%. Firstly, this study establishes a crude prevalence rate of T. gondii infection in cats. Secondly, it discusses the role of significant risk factors including sex, age and being either household or stray cats, in the epidemiology of the disease. Furthermore, the current study determines gaps and drawbacks in the prior studies that are useful to keep in mind to assist in designing more accurate investigations in future. of tissue cysts rather than oocysts. Surprisingly, a cat must ingest at least 1000 oocysts in order to develop an infection, although ingestion of just one bradyzoite is enough for a cat to acquire T. gondii infection (Dubey 2008).
Clinical manifestations of cats infected with T. gondii include depression, anorexia and fever, followed by peritoneal effusion, hypothermia, icterus and dyspnoea. Moreover, some other symptoms of toxoplasmosis are diarrhoea, weight loss, muscle hyperaesthesia, fever, anorexia, seizures, ataxia, pancreatitis and anterior or posterior uveitis (Dubey & Lappin 2006). Furthermore, the coccidian phase of the entero-epithelial cycle is seen solely in the definitive feline host. The extra-intestinal development of T. gondii is the same for all hosts, including all warmblooded vertebrates (Dubey 2005). The disease is important both in the medical and veterinary fields. Toxoplasmosis causes significant economic losses and damages to animal husbandry due to stillbirths and neonatal mortality in sheep and goats (Dubey & Beattie 1988;Hartley & Marshall 1957).
In spite of the need, currently there is no effective vaccine, even though many efforts have been conducted to develop a vaccine and are ongoing. Furthermore, no approved treatment exists for clinical toxoplasmosis in cats. Drugs including pyrimethamine, sulphonamides, trimethoprim and clindamycin, either alone or in combination, that have been prescribed to treat cats with clinical toxoplasmosis have shown varied results (Dabritz et al. 2007).
Cats have a key and crucial role in the epidemiology of toxoplasmosis, so expanding the basic knowledge about T. gondii infection in cats is a matter of importance. It is worth mentioning that epidemiological investigations are still the most useful method for evaluating the status of T. gondii infection. Despite the multitude of publications on toxoplasmosis in cats from Iran, there is no systematic review and meta-analysis that can describe the status of toxoplasmosis in the final host in this country. Therefore the objective of the current systematic review and metaanalysis was to determine the weighted prevalence of T. gondii infection and describe the epidemiological features of infection in cats in Iran.

Database search
To gather information a precise and comprehensive search was performed on all scientific publications (full texts and abstracts) from February to April in 2013 ( Figure 1). The following nine databases were included: five English databases (PubMed, Google Scholar, Ebsco, Science Direct and Scopus) and four Persian databases (Magiran, Irandoc, IranMedex and the Scientific Information Database [SID]). In addition to published articles, dissertations and all proceedings of national parasitology congresses held in Iran from 1975 to 2013 were carefully evaluated. In order to avoid missing any articles, whole references of papers were also meticulously checked.

Data collection
All cross-sectional studies carried out to estimate the prevalence of toxoplasmosis, diagnosed by different methods using serological, molecular and parasitological tests on cats, were included. Repetitive papers were excluded. The data that were collected for the current study were as follows: year of publication, first author, study areas, sample size, number of males and females, prevalence rate, age of samples, diagnostic tests, time of the study, and involving either domestic or stray cats. For this purpose a data extraction form was used.
The quality of the meta-analysis was evaluated using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist, which included 22 items that we considered essential for good reporting of observational studies. These items were related to the article's title, abstract, introduction, methods, results and discussion sections. Scores under 7.75 were considered to indicate bad quality, of 7.76-15.5 low quality, 15.6-23.5 moderate and more than 23.6 high quality (Von Elm et al. 2007).

Statistical methods
The crude and the weighted prevalence estimates as well as the 95% CI for each study that was included were calculated. A forest plot was used to visualise the heterogeneity amongst 12 studies from 11 papers were selected based on topics. 3425 papers from 9 databases (PubMed, Google Scholar, ISI web of science, Ebsco, Science Direct, Scopus, Magiran, Irandoc, IranMedex and SID) were idenƟfied.
Proceedings of Iranian parasitology congresses and theses were searched manually.
5 parasitology congress abstracts and 4 studies from 3 dissertaƟons of Iran universiƟes of medical and veterinary sciences were added.
Finally 21 studies which met our eligibility criteria were approved in the current review and meta-analysis. 3414 papers which did not deal with prevalence of toxoplasmosis in cats and duplicate papers were excluded. studies. The heterogeneity was expected in advance, and statistical methods such as I 2 and Cochrane's Q test (with a significance level of p < 0.1) were used to quantify variations. For the purpose of meta-analysis we assumed that the included studies were random samples from the populations under study, and a random-effect model was employed. Meta-regression and subgroup analyses were employed to assess the cause of heterogeneity amongst the selected studies, and Egger's regression test and funnel plotting were used to evaluate publication bias.
Proportions of individual studies and overall prevalence were presented using forest plots. The meta-analysis was performed using the trial version of StatsDirect statistical software (http://www.statsdirect.com).

Results
From the nine databases, 21 studies met the eligibility criteria and were included in the current systematic review and meta-analysis. A mean score of 16.8 using the STROBE checklist (Von Elm et al. 2007) was obtained for the 21 studies that were analysed.
A flowchart depicts the study design process (Figure 1). A total number of 2145 cats was examined for toxoplasmosis from 1975 to 2013 in different areas of Iran, and 662 cases were diagnosed as positive using different diagnostic methods (Tables 1 and 2).
During a period of 39 years, nine different types of diagnostic methods were employed to evaluate T. gondii infection in cats, as follows: the modified agglutination test (MAT), direct agglutination test (DAT), indirect immunofluorescent assay (IFA), latex agglutination test (LAT), immunochromatography test (ICT) and Sabin and Feldman Test (SFT), wet smear, flotation and polymerase chain reaction (PCR). The most frequently used diagnostic method for T. gondii assessment in cats in Iran was the IFA (5 studies), followed by flotation (4 studies), MAT (3 studies), LAT (2 studies), DAT (2 studies), ICT (2 studies), SFT, PCR and wet smear (1 study for each). The pooled proportion of toxoplasmosis, using the random-effect model, amongst cats in Iran over the 39-year period was estimated at 33.6% (95% CI 22.05-46.41) and a forest plot diagram of the current study was drawn (Figure 2). A wide variation was observed in the prevalence estimates of different studies (Q statistic = 742.3, df = 20, p < 0.0001 and I² = 99%).
The prevalence rate of cat toxoplasmosis in various regions of Iran was between 1.2% and 89.2% in Khorasan and Tehran respectively. The prevalence rate of toxoplasmosis in cats in different parts of Iran is shown in Figure 3. Amongst the studies included, only three studies compared stray cats with domestic cats for toxoplasmosis (Akhtardanesh et al. 2010;Haddadzadeh et al. 2006;Raeghi & Sedeghi 2011). They found a statistically significant difference in the number of positive stray cats and domestic cats. The results based on

ProporƟon meta-analysis plot [random effects]
0.0 0.2 0.4 0.6 0.8 1.0 age distribution were mentioned in 6 out of 21 surveys. A significantly higher prevalence rate of T. gondii was observed in older animals as compared with younger ones. Regarding sex, just two studies reported a significant difference, with male cats showing a higher prevalence rate of toxoplasmosis than females (Farhang 2010;Raeghi & Sedeghi 2011). The source of faeces in all the studies was the rectum of the examined cats.
Egger's regression test and a funnel plot were carried out for assessment of publication bias, and results showed that publication bias was significant ( Figure 6).
Five studies employed stool examination on 582 cats and 110 cases (18.9%) were reported as positive.

Discussion
The current study is the first systematic review and metaanalysis of cat toxoplasmosis in Iran. It provides valuable data about the prevalence of toxoplasmosis in cats from 1975 to 2013. The overall prevalence rate of toxoplasmosis amongst cats in Iran was estimated to be 33.6%. The worldwide seroprevalence of toxoplasmosis in domestic cats (Felis catus) was estimated to be 30% -40% (Dubey & Beattie 1988), and our findings roughly correspond to this range. Besides, previous studies have shown that the prevalence rates of T. gondii antibodies in cat populations vary greatly, Provinces with more than one study from 0% to 100%, and depend on particular criteria including age, method of survey, number of animals studied and geographical area (Dubey 2005;Dubey & Beattie 1988).
Several surveys have been performed on cat toxoplasmosis in Iran's neighbouring countries. An investigation on anti-T. gondii antibodies in 99 cats in Ankara, Turkey, using SFT and IFA showed prevalences of 40.3% and 34.3% respectively (Özkan et al. 2008), which is in close alignment with the present study. In another survey in Nigde, Turkey, the prevalence of antibodies to T. gondii employing SFT in stray cats was reported to be 76.4% (Karatepe et al. 2009). In Saudi Arabia the prevalence rate of T. gondii antibodies in cats was determined to be 15.2% using the indirect haemagglutination (IHA) test (Hossain et al. 1986).
A study on stray and domestic cats using LAT in Lahore, Pakistan, showed that 56% of cases were seropositive for T. gondii (domestic cats 48% and stray cats 64%) (Shahzad et al. 2006). In addition, the seroprevalence rate of T. gondii infection in cats was reported to be 60% in Faisalabad, Pakistan using LAT (Ahmad et al. 2001). The infection rates in cats in Lebanon and Iraq were reported to be 78.1% and 100%, respectively (Deeb, Sufan & Digiacomo 1985;Khairy, Alaa & Ahmad 2010). The findings of the current study are in agreement with those of Dubey (1998), who reported a 30% -80% seroprevalence rate in cats in the United States of America (USA).
A high prevalence of toxoplasmosis of cats in some areas may be due to the following factors: humid and temperate climate; absence of routine treatment for feline toxoplasmosis; and a considerable abundance of cats. It was confirmed that there is conformity between climate and the prevalence rate of toxoplasmosis, and regional prevalence varied in conformity with different climates. It is usually more prevalent in warm, humid climates and at lower altitudes compared to cold or dry districts. This fact is associated with longer viability of T. gondii oocysts in a warm and humid environment (Tutuncu et al. 2003). Sporulated oocysts of  T. gondii can persist in the environment (particularly in moist shaded soil or sand) for several months owing to their resistance to variable environ mental conditions (Dubey & Beattie 1988;Webster 2001).
In the USA the prevalence rate in the drier southwest, including New Mexico, Utah and Arizona, was lower (16.1%) than in humid climates such as Hawaii (59.2%), in accordance with the abovementioned climate template for T. gondii infection (Dubey & Jones 2008). In contrast, in some areas of Iran, such as Mazandaran province, the findings are not in accordance with that climate template.
Iran's climate ranges from arid or semi-arid to subtropical along the Caspian coast and in the northern forests. In the north of the country temperatures seldom decrease below freezing in winter, and the area usually remains humid for the rest of the year. In summer the temperature rarely Note: Weights are from random effects analysis.  (Ghorbani, 1983) Male (Raeghi and Sedeghi, 2011)   In the past different diagnostic tests were used in T. gondii surveys in cats. IFA was the most frequently employed test for the diagnosis of cat toxoplasmosis. This test was introduced in 1992 and considered a more reliable test than other serological diagnostic methods due to its significant sensitivity and specificity. This method is a relatively simple assay for evaluating the infection of animals, and also is particularly useful for screening a large number of specimens, which may explain why most studies used this method (Chejfec 1999;De La Luz Galvan-Ramirez et al. 2012;Silva et al. 2002).
In this systematic review and meta-analysis sex, age and type of cat (stray or domestic) can be probable causes of homogeneity, as well as publication bias. The role of risk factors including sex, age and being stray or domestic in the prevalence of toxoplasmosis is undeniable. In the current review there was a statistically significant difference between sexes (p < 0.05), and a higher infection rate of toxoplasmosis was seen in male cats compared to females. This may be related to lifestyle, as male cats have more of a tendency to wander and thus more access to contaminated sources; spending more time outdoors may increase their exposure to infection.
The age of animals is another major factor in prevalence of toxoplasmosis. Since young animals were less infected than older ones, it is expected that with increasing age exposure to T. gondii infection also increases (Miró et al. 2004). In Iran a significant relationship was observed between age at sampling and prevalence rate of toxoplasmosis amongst cats. These findings were in accordance with the results of some studies in other countries, which reported an increased prevalence of toxoplasmosis in older compared to younger cats (Frenkel et al. 1995;Maruyama et al. 2003;Vollaire, Radecki & Lappin 2005). It has been mentioned that newborn kittens are more dangerous than adult cats for transmission of this infectious disease, since they excrete oocysts for 1-2 weeks. On the other hand, kittens infected transplacentally or via milk exhibit more severe symptoms and frequently die due to pulmonary or hepatic disease (Buxton & Rodger 2008).
In the present study, Haddadzadeh et al. (2006), Akhtardanesh et al. (2010) and Raeghi and Sedeghi (2011) showed that the prevalence of toxoplasmosis in stray cats was noticeably higher than in domestic cats. In general, stray cats are more prone to T. gondii infection compared to household cats. The findings confirmed that stray cats have a tendency to have a higher prevalence rate than cats kept indoors. An acceptable justification for this fact might be that stray cats could acquire the infection through catching wild rodents, birds and reptiles, raw food scraps and so on. Cat owners can decrease their cats' exposure risk of acquiring T. gondii infection by keeping their cats indoors and also not feeding them uncooked meat and milk. In addition, the food source of the animal is significant in transmission and also for completion of the lifecycle of the parasite . Integrated control strategies and measures should be considered to prevent and control toxoplasmosis in both stray and household cats, which will have important implications for the control of toxoplasmosis in humans and other important intermediates such as sheep, goats and cattle.

Limitations of the study
The collected data are limited just to 10 out of 31 provinces, and there is a paucity of data for cat toxoplasmosis in the majority of provinces of Iran. Although many efforts have been made to determine the prevalence of toxoplasmosis in cats in different parts of Iran, some gaps in prior studies were evident.

Recommendations
Not enough attention was paid during sampling to the role of major factors including sex, age, and being stray or companion cats, despite their key role in the epidemiology of the disease. Therefore, considering all of the abovementioned parameters is necessary in order to overcome these shortcomings in future. The present study will form the basis of further studies that will enable us to deepen our knowledge of the epidemiology of T. gondii.

Conclusion
Based on the current results, stray cats are probably the major source of T. gondii infection in Iran. This study accentuates some valuable and interesting points: the prevalence rate of toxoplasmosis in cats in Iran is high (33.6%), and this considerable infection rate of final hosts can be considered a potential danger to public health and animals due to high contamination of the environment by oocysts.