Prevalence of Streptococcus suis serotype 2 isolated from pigs: A systematic review and meta-analysis

Background and Aim: Among Streptococcus suis serotypes, S. suis serotype 2 is the most significant serotype that causes serious diseases in pigs and humans worldwide. The present study aimed to estimate the global prevalence of S. suis serotype 2 isolated from pigs, determine its trend, and explore the factors associated with this serotype. Materials and Methods: We retrieved relevant published studies from PubMed, Scopus, and the Web of Science. The retrieved citations were screened for possible inclusion. Relevant data were then extracted from the included studies. The random-effects model was used for all meta-analyses. A subgroup meta-analysis was used to assess the heterogeneity of the prevalence for four characteristics (continents, sampling organs, reporting unit, and pig’s health status). A cumulative meta-analysis was performed to determine the cumulative prevalence over time. Meta-regression analysis was used to determine the trend of pooled prevalence of S. suis serotype 2 over time. Results: Of 600 articles retrieved, 36 studies comprising a total sample size of 6939 isolates or samples from 16 countries of four continents were included for meta-analysis. The pooled prevalence of S. suis serotype 2 isolated from pigs was 13.6% (95% confidence interval [CI], 10.7%–17.1%), with high heterogeneity among the included studies (Cochran’s Q, 431.6; p < 0.001; I2 = 91.9%; Table-1). No statistical significance was observed among subgroups of the four characteristics examined. However, the pooled prevalence of S. suis serotype 2 was as high as 16.0% (95% CI, 12.5%–20.3%; n = 16) in diseased pigs compared with 9.9% (95% CI, 5.6%–17.0%; n = 15) in healthy pigs. The pooled prevalence of S. suis serotype 2 isolated from pigs did not significantly decrease over time [regression coefficient = −0.020 (95% CI, 0.046–0.006, p = 0.139)]. The pooled prevalence of S. suis serotype 2 isolated from pigs fluctuated slightly between 13.2% and 17.8% from 2007 to 2023, although the pooled prevalence gradually decreased from 30.6% in 1987 to over 20% in 2003. Conclusion: The global prevalence of S. suis serotype 2 isolated from pigs was estimated to be 13.6% (approximately 10% in healthy pigs and around 16% in diseased pigs). S. suis serotype 2 isolated from pigs did not change significantly over time. These results indicate that S. suis serotype 2 remains a problem for the pig industry and poses a threat to human health.


Introduction
Streptococcus suis, which originated from pigs, is an emerging zoonotic pathogen that causes severe illnesses such as meningitis, deafness, septicemia, and even death in humans [1][2][3][4].Although S. suis is considered as a commensal bacterium residing mainly in the palatine tonsil and nasal cavity of pigs [5], some strains of S. suis can also cause severe diseases, especially in susceptible weaning pigs, resulting in a huge economic loss in the pig industry worldwide [6,7].In humans, the burden of S. suis infection is highly pronounced in Southeast Asian countries such as Thailand, Vietnam, and Indonesia [8,9].Almost all human cases in this region are associated with raw or undercooked pork or pork products [1,3,4].Several sociocultural factors, such as traditional culture, shared beliefs, socioeconomic level, and personal attitudes, play a role in human S. suis infection in Southeast Asian countries [10].In North America and Europe, people typically get infected through close contact with pigs or pork products; therefore, pig farmers and butchers are the occupations at risk [11].
S. suis is currently classified into 29 serotypes (1-19, 21, 23-25, 27-31, and 1/2) on the basis of Copyright: Keonam, et al.Open Access.This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.The Creative Commons Public Domain Dedication waiver (http:// creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Therefore, the study aimed to use a meta-analysis to estimate the global prevalence and trend of S. suis serotype 2 isolated from pigs and to explore some characteristics associated with the heterogeneity of the prevalence.

Ethical approval
This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (http:// www.prisma-statement.org/).Ethical approval was not required as no animals or animal-derived products were used in this study.

Study period and location
The literature search, data collection, and data analysis were conducted at the Faculty of Veterinary Medicine, Khon Kaen University, from June 2022 to July 2023.The included studies were published between 1987 and 2023.The included studies were conducted in 16 countries from Asia, Australia, Europe, and North America.

Search strategy
Two authors independently performed data search from PubMed, Scopus, and Web of Science to identify the relevant citations reporting the prevalence of S. suis serotype 2 isolated from pigs.The search was originally conducted in June 2022 and was updated in May 2023.The PECO structure [26] was used to formulate the research question (P, population = pigs or swine; E, exposure = positive case of S. suis infection; C, comparator = negative cases of S. suis infection; and O, outcome = the prevalence of S. suis isolated from pigs).The most important search keywords were "swine, Streptococcus suis, prevalence."For each database, the search was not limited to date but limited to English.The medical subject heading (MeSH) was checked for appropriate and complete keywords.An example search algorithm from the databases was follows: ("swine"[MeSH Terms] OR "swine"[All Fields] OR "swines"[All Fields] OR ("swine"[MeSH Terms] OR "swine"[All Fields] OR "pigs"[All Fields]) OR ("swine"[MeSH Terms] OR "swine"[All Fields] OR "suidae"[All Fields]) OR ("swine"[MeSH Terms] OR "swine"[All Fields] OR ("wart"[All Fields] AND "hogs"[All Fields]) OR "wart hogs"[All Fields]) OR ("swine"[MeSH Terms] OR "swine"[All Fields] OR "warthogs"[All Fields] OR "warthog"[All Fields])) AND "Streptococcus suis"[All Fields] AND ("epidemiology"[MeSH Subheading] OR "epidemiology"[All Fields] OR "prevalence"[All Fields] OR "prevalence"[MeSH Terms] OR "prevalance"[All Fields] OR "prevalences"[All Fields] OR "prevalences"[All Fields] OR "prevalent"[All Fields] OR "prevalently"[All Fields] OR "prevalents"[All Fields]).

Inclusion and exclusion criteria
Two independent reviewers carefully screened the titles and abstracts for initial inclusion after the duplicated citations were removed.Full-text articles that passed the first screening step were carefully examined under specific inclusion and exclusion criteria for final inclusion.This study was included if it reported the prevalence of S. suis serotype 2 isolated from pigs.The studies were excluded if (1) they were not related to the prevalence of S. suis serotype 2 isolated from pigs, (2) they were reviewed articles, (3) they were case reports, (4) they were experimental studies, and (5) they contained unclear information concerning the prevalence of S. suis serotype 2 isolated from pigs.Because S. suis serotype 2 is genetically very similar to S. suis serotype 1/2, conventional or multiplex polymerase chain reaction (PCR) cannot distinguish between these two serotypes.Therefore, agglutination and coagglutination tests using serotype-specific antisera or additional PCR techniques such as PCR-restriction fragment length polymorphism and whole-genome sequencing data are required for confirmation of S. suis serotype 2 [27][28][29].Therefore, studies reporting S. suis serotype 2 using PCR techniques without additional techniques for serotype confirmation, as mentioned above, were excluded from the study.To evaluate the effect of this exclusion, a sensitivity analysis was performed as described in the statistical analysis section.Disagreements regarding inclusion or exclusion were resolved by discussion.

Study quality assessment
A checklist for quality assessment adapted from the previous study that had five questions [30].These were as follows: (1) Was the research objective clearly described and stated?(2) Were the period and location of the study clearly stated?(3) Was the sample categorized into different subgroups?(4) Was the sampling method described in detail?(5) Were the diagnostic techniques clearly pointed out?Scoring of each question was based on a simple scale system ("2" for yes, "0" for no, or "1" for unsure); therefore, a total score for each included study was 10.

Data extraction
Two authors extracted and verified the data from the included studies.Disagreement regarding data extraction was resolved by discussion.Details of the extracted data were as follows: (1) study identification Available at www.veterinaryworld.org/Vol.17/February-2024/1.pdf (name of the first author and publication year), (2) study year, (3) study location, (4) pig's health status, (5) sampling organs, (6) number of total S. suis isolates or samples (sample size), (7) number of S. suis serotype 2 positive isolates or samples, and (8) detection methods.

Statistical analysis
Comprehensive meta-analysis program version 3 (Biostat, Englewood, NJ, USA) was used for all meta-analyses, including overall meta-analysis, subgroup meta-analysis, meta-regression analysis, cumulative meta-analysis, and assessment of publication bias.p < 0.05 was considered statistically significant for all analyses unless otherwise stated.To stabilize the variance of the raw prevalence before pooling, logit transformation was used as follows: logit(p) = ln(p/[1-p]), where "p" is the proportion and "ln" is the natural log [30].

Overall meta-analysis
A random-effects model was used to estimate the overall pooled prevalence and its 95% confidence interval (CI).Cochran's Q and I 2 statistics were used to assess the heterogeneity.A Cochran's Q test p < 0.05 implied significant heterogeneity.I 2 values of 25%, 50%, and 75% were considered low, moderate, and high heterogeneity, respectively [31].Individual studies were used as a unit of analysis for the overall pooled prevalence.

Subgroup meta-analysis
A subgroup meta-analysis was used to assess the heterogeneity of the pooled prevalence of S. suis serotype 2 isolated from pre-defined pig characteristics.Subgroup meta-analysis was performed for the following four characteristics: (1) location of the study (among continents), (2) sampling organs (tonsils versus other organs), (3) unit of the report (isolates versus samples), and (4) pig's health status (healthy versus diseased).For all subgroup meta-analyses, the subgroup within the study was used as a unit of analysis.

Meta-regression and cumulative meta-analysis
Meta-regression analysis was used to determine the pooled prevalence trend of S. suis serotype 2 isolated from pigs.Meta-regression of the pooled prevalence was performed against the study year.A median of multiple years was used for analysis in a study reporting a multiple-year study.The expected study year was used in a study that did not report the study year.We calculated the expected study year by subtracting 2 years from the publication year (median difference between the publication year and study year of the included studies).A cumulative meta-analysis was also used to determine the cumulative evidence of the pooled prevalence of S. suis serotype 2 isolated from pigs over time (the publication year).

Sensitivity analysis
Sensitivity analysis was conducted to assess the robustness of the pooled prevalence estimation of S. suis serotype 2 infections in pigs.First, we compared the results of the fixed-effects model with those of the random-effects model.Second, we compared the results of using a study as a unit of analysis with those of using a subgroup as a unit of analysis.Third, we compared the results of the included studies with those of the excluded studies reporting S. suis serotype 2 but did not describe additional serotype confirmation techniques.Finally, we conducted a leave-one-out meta-analysis to evaluate whether each individually included study influenced the results.

Assessment of publication bias
Publication bias was assessed using Begg's test and Egger's test with p < 0.1, indicating the presence of publication bias [32,33].Publication bias was also visually assessed using a funnel plot.Duval and Tweedie's trim-and-fill method [34] was used to estimate the adjusted prevalence of S. suis serotype 2 isolated from pigs in the case of asymmetrical funnel plots.

Characteristics of the included studies
Table-1 [15][16][17][18][19][20][21][22][23][24][25] summarizes the characteristics of the 36 studies included in the study.Of the 36 included studies, 6939 isolates or samples were examined, resulting in 962 isolates or samples positive for S. suis serotype 2. The median sample size was 118 isolates (range, 15-680 isolates).Data were reported from 16 countries from four continents (12 studies from Asia, one study from Australia, 14 studies from Europe, and nine studies from North America) (five studies each from Canada, Spain, and Thailand, four studies from China and the United States, and 13 studies from other countries).Of the 6939 isolates or samples examined, 2032 were from Asia, 2617 from Europe, and 1749 from North America.Of the 6939 isolates or samples examined, 2388 were from healthy pigs, 2973 were from diseased pigs, and 1578 were from pigs with mixed or unspecified health status.

Study quality assessment
This study's quality assessment tool was based on the full 10-point rating scale.The mean ± standard deviation of the overall quality scores of all included studies was 9.03 ± 1.38.Regarding the range of scores

Meta-regression and cumulative meta-analysis
Meta-regression analysis showed that the global prevalence of S. suis serotype 2 isolated from pigs did not significantly decrease over time [regression coefficient = −0.020(95% CI, −0.046-0.006,p = 0.139)].The meta-regression equation is expressed as the logit event rate = 37.49−0.020(year), where year is the study year (Figure -3).

Table-2:
Study quality assessment showing the number of the included studies in each category of the simple rating scale based on a checklist of five items [30].

Yes No Unsure
Was the research objective clearly described and stated?Was the sampling method described in detail?(Figure -5).The trim-and-fill method revealed nine missing (hypothetical) studies.The adjusted point estimate of the overall pooled prevalence of S. suis serotype 2 isolated from pigs somewhat increased from 13.6% (95% CI, 10.7%-17.1%) to 18.3% (95% CI, 14.5%-22.8%)after these nine studies were imputed in the model.

Discussion
Our study synthesized the pooled prevalence of S. suis serotype 2 isolated from pigs from 36 included studies worldwide with a total sample size of 6939 isolates or samples.The overall pooled prevalence of S. suis serotype 2 isolated from pigs in our study was estimated to be 13.6% (95% CI, 10.7%-17.1%)with high heterogeneity.The point estimate of the pooled prevalence in our study is quite high, indicating that pigs remain a potential source of S. suis serotype 2 contamination.The high heterogeneity of the result was within our expectations.Therefore, subgroup meta-analysis and meta-regression were used to assess the heterogeneity or variability of the prevalence among the included studies.Below is our detailed discussion.
Our subgroup meta-analysis did not find a significant difference among the three continents (Asia, Europe, and North America; Australia was excluded from the analysis because it had only one study).The pooled prevalence of S. suis serotype 2 isolated from pigs did not significantly differ among continents, ranging from 12.2% (95% CI, 7.8%-18.6%) in Asia, 14.1% (95% CI, 8.8%-21.7%) in Europe, and to 16.1% (95% CI, 11.1%-22.8%) in North America.For all continents, the 95% CI of the point estimate was very wide.This indicated that there was a high variability in the prevalence data from the included studies within each continent.In Asia, the prevalence of S. suis serotype 2 in pigs was as low as 2.2% in Thailand [53] and as high as 64.5% in China [66].In Europe, the prevalence of S. suis serotype 2 was as low as 0% [46] in Belgium and as high as 50.6% [60] in Spain.The prevalence of S. suis serotype 2 in the United States was as low as 0% in the study of Amass et al. [45] and as high as 39.4% in the study of Mogollon et al. [54] in North America.
This phenomenon has also been observed in other countries, such as Thailand [20,53], Spain [58,60], and Canada [15,52].The great variability in the prevalence of S. suis serotype 2 isolated from pigs among the included studies, even in the same region, reflected the difference in study settings of the included studies and depended on several factors such as coinfection with porcine reproductive and respiratory syndrome virus at weaning, introduction of pigs into the nursery, sow parity, relative humidity, and temperature [7,70,71].
Although our study showed that the pooled prevalence of S. suis serotype 2 was not significantly different between diseased pigs and healthy pigs, the pooled prevalence was as high as 16.0% in diseased pigs compared with 9.9% in healthy pigs (i.e., 1.6 times higher in diseased pigs).The non-significant result may arise from the small sample size within the subgroup resulting in a low statistical power.Our Available at www.veterinaryworld.org/Vol.17/February-2024/1.pdf Specific methods or techniques that used to differentiate S. suis serotype 2 from serotype 1/2 were clearly described (confirmed) or (unconfirmed).The unconfirmed 10 studies were excluded from the data synthesis.b Removed the study of Wang et al., [66], c Removed the study of Marois et al., [51], CI=Confidence interval results indicated that S. suis serotype 2 was associated with diseased pigs because the prevalence of S. suis serotype 2 in diseased pigs was 1.6 times higher than that in healthy pigs.S. suis serotype 2 is known as a virulent strain that causes diseases in humans and pigs worldwide [11,13].Clinical manifestations in pigs include septicemia, meningitis, endocarditis, pneumonia, and even death [5,11].This impacted the health, welfare, and economic loss of pigs [6].Most S. suis outbreaks in humans result from S. suis serotype 2 [1,11].Important risk factors for human infections include raw pork consumption, exposure to pigs or pork, pig-related occupation, and male sex [8].The high pooled prevalence of S. suis serotype 2 (9.9%, 95% CI, 5.6%-17.0%;n = 15 subgroups) in healthy pigs was surprising.A high prevalence of S. suis serotype 2 in healthy pigs may be a potential source of infection for other pigs and humans [5,72].Horizontal transmission primarily occurs through the oro-nasal route or nose-to-nose contact in pigs [5].In addition, many of the included studies used samples from healthy pigs or pork products from slaughterhouses or wet markets [20,49,53,55,65,67].This production step is close to-and represents a risk for consumers.
For ease of comparison, we divided the sampling organs into two main categories (tonsils vs. other organs).The pooled prevalence of S. suis serotype 2 in pigs was not significantly lower in tonsils than in other organs (13.2% and 14.4%, respectively).The palatine tonsil of a pig is a primary organ for the early colonization of S. suis [5].S. suis can be found in other organs, such as the lung, brain, spleen, blood, joint fluid, saliva, oral swab, nasal swab, vaginal swab, tongue swab, and pleural effusion [25,53].Samples from other organs mainly come from studies of diseased pigs [15,20,21,23,24,27,52,62]. Because organ collection sampling was confounded by the pig's health status, further analysis was performed to reveal this confounding effect.Diseased pigs and healthy pigs were separately analyzed on the basis of the sampling organs.The pooled prevalence of S. suis serotype 2 was 28.6% (95% CI, 14.8%-48.0%,n = 4) in the tonsil samples and 13.9% (95% CI, 11.0%-17.4%,n = 12) in the other organ samples in diseased pigs.In healthy pigs, the pooled prevalence of S. suis serotype 2 was 8.2% (95% CI, 4.1%-15.6%,n = 10) in the tonsil samples and 14.4% (95% CI, 4.0%-40.4%,n = 5) in the other organ samples.These results reveal a confounding relationship between the sampling organs and the pig's health status.
Most of the included studies reported isolates as a unit of report for the prevalence (n = 33 subgroups) compared with samples as a unit of report for the prevalence (n = 7 subgroups).The pooled prevalence of S. suis serotype 2 was not significantly different between isolates and samples (13.6% and 19.7%, respectively).Most of the included studies used isolates as a unit of report because they usually involved identifying all S. suis isolate serotype.Therefore, it is more convenient to report each serotype as a proportion of the total isolate.Some included studies [20,45,47] reported the prevalence of S. suis  serotype 2 directly from a positive sample out of a total number of positive samples.A non-significant difference in the pooled prevalence of S. suis serotype 2 in the reporting unit indicates that both reporting units are satisfactory.
No significant difference among subgroups in our subgroup analysis does not mean that there is no heterogeneity among subgroups.The number (n) of subgroups in our analysis was small due to the small number of studies included.This may result in a low statistical power and result in a false negative error.If more data are available, we will be able to prove the existence of significant subgroup differences.For example, a significant difference in the pooled prevalence of S. suis serotype 2 between diseased pigs and healthy pigs (16.6% [95% CI, 13.2%-20.8%]vs. 9.6% [95% CI, 5.8%-15.4%],respectively) was found if we added ten more studies (that we excluded) reporting S. suis serotype 2 but without describing additional techniques for serotype confirmation in our analysis.
Our meta-regression analysis showed that the pooled prevalence of S. suis serotype 2 in pigs did not change significantly over time.Our cumulative evidence showed that the pooled prevalence of S. suis serotype 2 in pigs fluctuated slightly from 13.2% to 17.8% between 2007 and 2023.Both results indicate that S. suis serotype 2 remains a source of problems for pigs and poses a threat to human health.In addition, S. suis rapid antimicrobial resistance may exacerbate the problems [73][74][75].
Sensitivity analysis was used to check the robustness of the decision-making results during the systematic review and meta-analysis.Our sensitivity analysis indicates that the point estimate of the pooled prevalence of S. suis serotype 2 in pigs is robust for all categories except the model of choice used in the analysis.
The point estimate of the random-effects model (the model of choice of our study) was 13.6%.It was somewhat lower when compared with the fixed-effects model (16.9%).However, the random-effects model was more appropriate in our study because the study settings, sampling designs, and other conditions of the included studies varied greatly [76].The results of the leave-one-out analysis indicate that individual studies do not distort the point estimate of the pooled prevalence of S. suis serotype 2 isolated from pigs.When we removed the study of Wang et al. [66] and that of Marois et al. [51], the point estimate was the lowest (12.9%) and the highest (14.1%), respectively, from the analysis.These values were very close to the point estimate from all the included studies (13.5%).
Regarding publication bias, Egger's test (p = 0.027) and Begg's test (p = 0.072) indicated publication bias in our study.A funnel plot also showed an asymmetrical distribution of the included studies.In addition, the trim-and-fill method revealed nine missing (hypothetical) studies.After these nine hypothetical studies were imputed in the model, the point estimate of the overall pooled prevalence of S. suis serotype 2 isolated from pigs increased from 13.6% (95% CI, 10.7%-17.1%) to 18.3% (95% CI, 14.5%-22.8%).This indicates that the pooled prevalence of S. suis serotype 2 isolated from pigs would be higher than we estimated without publication bias.
Our study has some limitations.First, our study focused only on the pooled prevalence of S. suis serotype 2 because this serotype is very important in human disease.Therefore, the pooled prevalence of all S. suis serotypes would be much higher because S. suis has 29 specific serotypes [13].A recent study estimated that the pooled prevalence of S. suis in pigs in China during 2000-2021 was as high as 40% [77].This means that S. suis poses a potential threat to the health of pigs and humans.Second, the study locations were mainly from Asia, Europe, and North America.There was only one study from Australia and no study from Africa.Therefore, subgroup analysis for Australia and Africa is not feasible.Interpretation of the global prevalence and subgroup analysis of S. suis serotype 2 with regard to the study locations should be carefully performed.Additional information from other regions (if available) may affect the point estimate of the pooled prevalence.Third, we included only studies published in English and retrieved from three databases (PubMed, Scopus, and Web of Science) due to limitations in our research team.Therefore, we may have missed data from other languages, which may have changed the point estimate of the pooled prevalence.

Conclusion
A random-effects meta-analysis was used to estimate the pooled prevalence of S. suis serotype 2 isolated from pigs from 36 included studies containing a total of 6939 isolates or samples from 16 countries of four continents.The overall pooled prevalence of S. suis serotype 2 isolated from pigs was 13.6% (95% CI, 10.7%-17.1%)with high heterogeneity among the included studies.The point estimate may be lower than the real-world situation due to the presence of publication bias.The pooled prevalence of S. suis serotype 2 was as high as 16.0% in diseased pigs and 9.9% in healthy pigs.There was no significant change in the pooled prevalence of S. suis serotype 2 isolated from pigs over time.The results of this study indicate that S. suis serotype 2 remains a threat to pigs and humans worldwide.
from 5 to 10, the median score was 9.All study quality assessment results are presented in Table-2.

Figure- 2 :
Figure-2:The prevalence estimates of Streptococcus suis serotype 2 isolated from pigs from the most reported countries.

Figure- 3 :
Figure-3: Scatter plot showing results of meta-regression analysis for the correlation between logit event rate (Y-axis) and study year (X-axis).A circle represents each included study (n = 36).

Figure- 4 :
Figure-4: Forest plot of cumulative prevalence and 95% CI of S. suis serotype 2 isolated from pigs over time.From 2007 to 2023, the pooled prevalence of S. suis serotype 2 isolated from pigs fluctuated between 13.2% and 17.8%.

Figure- 5 :
Figure-5: Funnel plot for visual assessment of publication bias.X-axis is the effect size (the logit of the event rate).Y-axis is the precision (standard error).A blue circle represents each included study.A red circle represents an imputed study (or missing study) to make the funnel plot symmetry.

Table - 1
: Characteristics of the included studies.

Table - 3
: The overall pooled prevalence of S. suis serotype 2 isolated from pigs and subgroup meta-analysis.
a Two subgroups were removed from analysis because low number of sample (Australia n = 1, Mixed [Asia-Europe] n = 1).b One subgroup was removed from analysis due to an unspecified sampling organ.c Nine subgroups were removed from analysis due to mixed or unspecified health status, CI=Confidence interval

Table - 4
: Sensitivity analysis to assess the robustness of the result estimates.